VOLUME FOUR
PORE PRESSURE PREDICTION
1. INTRODUCTION
Probing the earth’s subsurface for oil and gas presents many challenges and surprises.
Developing an understanding of this subsurface and attempting to predict, with
reasonable success, what lies ahead is a major significant factor in drilling safely, drilling
economically and drilling useable, productive wells. At the core of this understanding
should lie a strong fundamental knowledge of pore pressure, it’s development, anomalies
associated with normal, abnormal, and subnormal pore pressure and predictive
techniques which can be used as well planning and real time drilling tools.
Certainly, it is true that not all wells drilled world-wide are planned or programmed based
upon pore pressure predictions. However, this does not eliminate the need for
knowledge in this area since drilling environments are constantly changing and, even
though abnormal pressure may not be present, normal or subnormal pressures may be.
Prediction, evaluation and reaction to these environments is necessary (Figure 4A.1).
Figure 4A.1
WHY?
• Drill Fast
• Drill Safe
This introduction presents current technology, equations, and some examples of pore
pressure prediction techniques. It should be kept in mind that the material presented
here does have its limitations, but when consistently and carefully applied, it is a very
useful tool, from both a well planning standpoint and a “real time” drilling standpoint.
Two key points are worth mentioning at this time. First, the accuracy of these techniques
and the usefulness of the results are directly proportional to the amount of historical and
offset information used. Secondly, as drilling engineers, a great deal of our success and
well planners will stem from our ability to communicate with local exploration staffs and
obtain as much information as possible. As drilling engineers, we must be aware of the
data and information needed and be able to communicate this to the exploration
geologist.
2. SEDIMENTATION
Thousands of feet of sediment have been deposited over millions of years. It started as
soon as the earth had cooled enough to allow rainfall and has continued until today.
Historical geology is a fascinating study and makes excellent reading for any drilling
engineer.
Consider the drainage area of the Mississippi River. From Jackson Hole, Wyoming,
comes pieces of stone that are deposited south of New Orleans as sand. From Ely,
Minnesota, comes pine needles, leaves and more sand, and, further down the river, silt,
grass and other organic material. Reason suggests that more silt comes down the river
now than did when the drainage area was covered by grass and trees.
As this material reaches the Gulf, the sand settles out first near the shore. In deeper
water, only mud, silt, and organic material reach the ocean floor. The depth of abnormal
pressure can be a function of distance from a major river during the depositional phase.
3. COMPACTION
Consider one cubic foot of sediment just settled to the ocean floor in the Gulf of Mexico.
Just deposited, it would be hard to tell mud from water, but as it rests on bottom, the
solid material would settle to bottom and the water would flow away. Finally, one cubic
foot of mud is left (Figure 4A.2).
To assist in the analysis of this condition, consider the soil boring analysis in figures 4A.3
3
and 4A.4. Examine sample #1. It is interesting to note that its density was 89 lb/ft , or
11.9 ppg. This is certainly not yet shale.
We believe the specific gravity of normally-compacted shale to be about 2.6 (21.7 ppg).
Although we do not have shale yet (sample #1), we might assume that the grain density
of the sediments is 19.0 ppg. Also, assume that the density of the sea water is 9.0 ppg.
We can now analyze this first foot of soil to determine its grain content and water
content.
The interesting thing about compaction is that each cubic foot of mud below this has no
way of getting rid of its water with the exception that it leak through the cubic foot we are
considering. So, as this cubic foot compacts under weight, it gives off water but receives
more water from below. Thus, compaction is a very lengthy process.
This newly deposited cubic foot of mud also contains organic material that will give off
methane gas, further aggravating the process of compaction, and certainly complicating
the drilling process.
To illustrate this point further, consider the last data point on the referenced report. Note
that at a penetration depth of 698 feet, the density was 122 lb/ft3 or 16.3 ppg.
Proceeding with a similar analysis, we find the following:
This analysis indicates that even at 698 feet, grain-to-grain contact has not yet been
established and we certainly do not yet have shale. It is worth noting that in many young
sedimentary basins, this grain-to-grain contact is not established until a depth of possibly
3000 - 5000 feet, as shown in Figure 4A.6.
One final observation concerning this soil boring report is worth noting. Consider the plot
of density versus depth on semi-logarithmic scale as shown in Figure 4A.5. The first 250
- 270 feet below the sea floor seems to be compacting at a different rate than those
sediments below. Actually, the top sediments are moving and very unstable. This, of
course, contributes to a very difficult drilling environment.
Figure 4A.7
Compaction
• Near the surface, sediments act partially like rock and partially
like mud.
4. NORMAL PRESSURE
The process by which mud is changed into a solid as sedimentation occurs is called
compaction. Generally in the Gulf of Mexico the sediments do not achieve grain-to-grain
contact until a depth of 3,000 - 5,000 feet has been achieved. Typically in hard rock
environments, like West Texas, the unconsolidated interval may only be 100 - 200 feet,
obviously a much different environment. The fundamental point however, is that when
grain-to-grain contact has been established, and the water in the rock is free to move,
normal pore pressure exists.
This normal pressure is dependent on two parameters: 1) Pore fluid density, and 2)
Vertical fluid column height, as shown in Figure 4A.8. For most young sedimentary
drilling environments, the fluid density in the rock pore spaces will be about 9 ppg, or
exhibit a pressure gradient of .468 psi/ft. This is somewhat different in older hard rock
environments where the formation waters may be less saline (lower density) and the
water table may be lower. It is not uncommon to find effective fluid densities at depth to
be as low as 8.25 ppg (.429 psi/ft). Note that this is "effective” density and indeed can be
less than fresh water. To summarize, normal formation pressure is simply the
hydrostatic pressure exerted by a continuous fluid column at some depth, as in Figure
4A.9. It is dependent on fluid density and the vertical column height of the fluid.
common to find that a straight line can be drawn through formation density only after
grain-to-grain contact has been established and covering only those sediments which are
normally pressured. Therefore, it could be said that normal pressure exists when
formation density increases with depth in such a way that a straight line can be drawn
through the plotted points on semi-log paper. This straight line is called the "normal trend
line" and the slope of the line is an indicator of the rate at which the shale has
compacted.
To summarize, the key points about normally pressured sediments are (Figure 4A.10):
Figure 4A.10
Normal Pressure
Finally, it is well to note that even though grain-to-grain contact exists and the fluid is not
supporting any of the weight above it (overburden), compaction is still continuing,
porosity is being reduced and density is therefore, increasing (Figure 4A.11).
5. ABNORMAL PRESSURE
A barrier or seal can develop relatively suddenly as the result of tectonic movement of
the earth, or can develop slowly, in the form of a cap rock resulting from fluid and debris
movement through long shale segments. Long shale intervals will almost always contain
abnormal pressure. There are two reasons why this occurs. First, in order for fluid to
move from the inner section of any rock, a differential pressure must exist. Thus shale
internal pressure must be greater than its external or adjacent sand pressure. This is
provided that the fluid is moving.
Second, as the shale is compacting under the overburden load and fluid is moving out,
eventually the moving fluid, containing fine debris particles will begin to plug the
decreasing porosity channels in the shale. As this plugging effect develops, less and
less water will escape, a very dense or hard spot will develop, and fluid flow will slow
dramatically or stop. The end result is that the pore space fluid is supporting part of the
overburden. Abnormal pore pressure has, therefore, developed (see Figure 4A.13).
Encountering a very hard spot when drilling long shale sequences may very well be a
forewarning of a "drilling break" caused by abnormal pressure. This may not cause a
flowing well problem but can, depending on the degree of differential pressure, cause
shale sloughing or severe wellbore stability problems. Under these conditions, increasing
the drilling fluid density may very well cure the sloughing or stability problems. Figure
4A.14 is a graphical indication of what a depth versus density plot may look like under
these conditions. Note the existence of the cap rock (seal) as indicated by a hard spot
directly above an area which is less dense and, therefore, under-compacted. Abnormal
pressures range in magnitude from any pressure exceeding the normal gradient (8.25 -
9.00 ppg) to approximately a pretrostatic gradient (1 psi/ft or 19.23 ppg), Figure 4A.15.
Formation pressures in excess of 18.0 ppg are seldom encountered but certainly do
exist. in some areas, pressures in excess of what is considered to be a normal
overburden, have been recorded. These are extremely rare but also do exist. Some of
the typical causes of abnormal pressure are illustrated in Figures 4A.16, 4A.17 and
4A.18.
Years ago the main indicator of abnormal pressure was a kick or even a blowout.
Increasing the drilling fluid density seemed to be the answer to prevent these
catastrophes. It was soon discovered, however, that by indiscriminately increasing the
fluid density other problems arose. Namely, lost circulation, stuck pipe, and even
additional wellbore kicks and blowouts. Obviously, it is most desirable to drill with the
fluid density as close as possible to the formation pressure. Above all, we must drill
safely, but at the same time, drill efficiently with minimum wellbore and fluid problems.
Two related concepts had to be developed in order that this might be accomplished.
First, the ability to predict formation pressures had to be developed and secondly,
methods or indicators of abnormal pressure while drilling had to be recognized and
understood.
There are three stages in pore pressure determination 1) Before, 2) During, and 3) After
(Figure 4A.19).
Figure 4A.19
Before refers to prediction. We will
Stages of Detection first discuss several methods for
predicting the existence and
magnitude of abnormal pressures. In
• Before relatively young sedimentary basins,
shale property trends can be used
• During very effectively to illustrate how rock
density varies with depth and
therefore, can also be used to predict
• After pore pressures.
The reasons why shale trends are used are straight forward but certainly worthy of
mentioning. Generally, shale is composed of fine organic and mineral substances of
more or less uniform particle size. But more importantly, shale compacts uniformly and
predictably. Thus, shale sequences do have a “normal” trend line illustrating an increase
intensity with depth, as depicted in Figure 4A.20.
The choice of sand as a predictive medium fails primarily because sand can mean many
different substances and rock configurations. As an example, a sand in South Louisiana
is loose and unconsolidated, whereas a sand in West Texas may have grains that have
been cemented together, with the resulting rock as hard or harder than limestone.
Further complicating the picture is the fact that sand can be any grain size from gravel to
fine silicon silt. Sand, therefore, does not exhibit any sort of normal trend and cannot be
used as a predictive medium.
Shale property trends are by far the best indicators of abnormal pressure. Shale density
does increase with burial depth and the rate at which this increase is occurring is
represented by the slope of the normal trend line. Any departure from the normal
compaction trendline, indicating a less dense region, also indicates an under-compacted
and possibly over-pressured area. This departure from the normal compaction line is
called the “transition" zone.
From a well planning standpoint, it seems logical that, after developing a sound
subsurface correlation between a new well to be drilled and previously drilled wells, any
and all data should be used to develop shale property trends so as to indicate where
transition zones can be anticipated and what the magnitude of the abnormal pressure
might be. This data can also be used to calculate an anticipated fracture gradient. This
is a very important element in the well planning and drilling process and will be discussed
later.
We will now examine some of the “tools", listed in Figure 4A.21, used by the well planner
to determine the possible existence or non-existence of abnormal pressure. It is
important to note that effective and accurate well planning can only be accomplished
when a wealth of information is acquired and used. We can very seldom do a cost
effective job of well planning if we have only one set of data.
Figure 4A.21
Figure 4A.22
Before Drilling
Offset Well Logs
• Offset Well Logs
• Mud and Bit Records OFFSET WELL LOGS • Density
• Drilling Reports (Figure 4A.22): Good
• Resistivity
• offset logs are probably
the best source of positive • Conductivity
Geological Modeling
• Seismic data we have. If utilized • Sonic
• ROP Plots properly they are the best
• ITT instrument for establishing shale property trends
and therefore, are excellent indicators of
transition zones and instruments for quantifying pore pressure. We will confine our
discussion to the following parameters as they are the most commonly used.
Fundamental to understanding these parameters and how they relate to pore pressure
within the shales, is the fact that we correlate each of these with porosity or the degree
to which the shale has been compacted. As was stated earlier, shale does compact
uniformly and predictably. This suggests that under normal compaction, porosity
decreases uniformly and predictably.
Obviously, if under-compacted shale is encountered, its porosity will be greater than what
is above it in the geological sequence. This is truly an anomaly and can be seen in each
of the log derived parameters listed above. Considering shale density and normal
compaction, porosity will decrease with depth. Thus, it follows that density will increase
with depth for normal compaction. Should under-compaction be encountered the density
will decrease with the increase in porosity. Thus, plotting density versus depth will
establish a "normal" trend line. Deviation from this line with density showing a
decreasing trend will indicate under-compaction and possibly abnormal pressure. See
Figure 4A.6 for an example plot.
Second, consider the effects on shale resistivity. For normal compaction, with porosity
decreasing and density increasing, shale resistivity will be increasing. Understand that
as compaction is taking place, water is being forced up and out of these sediments.
Water is the conductive medium, therefore, conductivity must be declining and resistivity
must be increasing. The opposite trend occurs when drilling a transition zone. When
under-compaction is present, porosity has increased, density has decreased, and
resistivity has, therefore, decreased.
Again, this is due to the presence of more water, therefore, a more conductive, less
resistive rock. By considering resistivity, we have in fact also considered conductivity.
Without too much redundancy, it is sufficient to say that under normal compaction, water
is being driven out, therefore, conductivity must be decreasing. Upon entering an under-
compacted region with an increased porosity and water content, the conductivity will be
increased.
Finally, the sonic log must be considered. The sonic log actually indicates the interval
transit time (T), of a sound wave traveling through the formation and back to a receiver.
The units indicated on the sonic log are micro-seconds per foot (∆sec/ft). Note that sonic
velocity (feet/sec) is simply the reciprocal of the interval transit time multiplied by 106 (Vel
= 106 / T).
Sonic log analysis for pore pressure prediction is developed around the concept that as
porosity decreases and density increases, for normal compaction with depth, the rock
becomes a much more efficient sonic conductor. The sonic velocity will increase with
depth for normal compaction. Thus travel time (T) will decrease with depth, if
compaction is uniform and considered normal. It follows that, when under-compaction
exists, the sonic velocity will decrease thereby indicating an increasing interval transit
time.
One final parameter which should be mentioned here is temperature. The earth's core is
obviously hotter than its surface, therefore,. heat moves from the center to the surface.
This phenomena creates a temperature gradient which is generally between 1°F and
2°F per 100 ft. The earth's sediments are actually functioning as a heat exchanger and
the flow rate of heat through any formation is directly proportional to the formation
density. The higher the formation density, the smaller the temperature drop required to
generate a given heat flow.
Since abnormally pressured sediments are generally less dense than the normally
pressured sediments above, there is generally a measurable increase in flow line
temperature if abnormal pressure is encountered. A plot of differential temperature per
100 ft versus depth will be a straight line through normally pressured sediments. The
slope of that line will be in the range 1°F to 2°F per 100 ft. Upon drilling abnormally
pressured sediments (less dense formations) the plot of differential temperature per 100
ft. will show an increasing slope which is indicative of the earth functioning as a less
efficient heat exchanger.
More temperature drop is require to maintain a given heat flow rate. A plot of this type
would look very similar to an interval transit time versus depth plot, or would correlate
positively with a plot of formation density versus depth (Figure 4A.24). Flow line
temperature is another indicator of abnormal pressure.
OTHER USEFUL WELL DATA: All available offset well information and data should be
employed when developing any well program. Reliable drilling reports, drilling fluid
recaps, bit records, geological information and seismic data can all be used to enhance
the accuracy and reduce risk factors when developing a well plan.
Any information which may be used to determine transition zones, or to qualify formation
pressure is extremely valuable (Figure 4A.25). Any pressure data which can be
stratagraphically correlated to the well being planned always provides a point of known
pressure, which may be needed to establish a complete pore pressure plot for the well.
Drilling fluid recaps and bit records can also provide important information (Figure 4A.26).
Figure 4A.26
Figure 4A.25
Other Useful Data
Drilling Reports
Fluid Recap
• Mud Logs • Lost Circulation Zone
• Fill on Trip • Stuck Pipe Occurrences
• Torque & Drag • Kick Information
• “d” Exponent Plots
Bit Records
• Bit Type (Insert, Mill Tooth, PDC)
• Formation Drillability (Density)
Bit records may also indicate valuable drilling information, and in some situations may
actually provide the data necessary to quantify formation drillability. This relates to
formation density when evaluating shales. With sand-shale sequences, formation
drillability can be quantified using the “d” or "dc” exponent concept. This will be
discussed in detail later, however, since the “dc” exponent responds to formation
drillability, it can and often is used to quantity pore pressure and is very useful in
determining transition zones. Some problems may occur if the "dc" exponent is used as
the sole tool for predicting pore pressures and must be understood. These pitfalls will be
outlined in a later section.
Geological information can help determine the location of faults and the depositional
environment of the formations (Figure 4A.27). For example, nearly all anticlinal reservoirs
are broken by faults. Usually they are vertical and strike at an angle of about 70° to the
axis of the anticline. The depositional environment affects permeabilities and drillabilities.
Figure 4A.28 illustrates the electric log response for several depositional environments.
The alluvial fan and braided-stream deposits show as stacks of sand with thin shale
beds. The point bars nearly always show the abrupt base and narrow top (bell shape),
while the stream-mouth and barrier bars show the broad, abrupt top and gradational base
(funnel shape). The turbidities show stacked sand bodies separated by shale beds.
Figure 4A.29 shows the electric log response of beach deposits. The log response is the
inverse of that for stream channel sands.
Beach sands are deposited upon fine-grained sediments that have little porosity and
reduced SP and resistivity response. Correspondingly, in a beach environment
permeabilities decrease from top to bottom.
Seismic data usefulness is shown in Figure 4A.30 which depicts a seismic section for a
growth fault of increasing angle below which the reflections appear very broken. Some of
these featureless shale zones may be caused by diapirism deep below the surface, while
others may represent the toe zone of the slump block where the fault emerges at the
surface. Shale in this chaotic zone is under-compacted and contains fluids at pressures
almost equal to the weight of the overburden. When the pressure in the pore water
approaches the weight of the overburden, the overlying strata are practically floating.
The weight of the overburden (S) is sustained by the stress in the skeleton of the solid
grains (σ) and the pore pressure (p) in the interstitial fluids. (Figure 4A.31)
S = + p
As p increases, decreases and may become very small. That is, the solid skeleton is
supporting very little weight, and the overlying strata are floating. Thus, they can slide
under weak lateral forces, such as gravity sliding if the area is tectonically tilted. Most, it
not all, low-angle thrust faults probably take place in a zone of abnormally high pressure.
Rate of penetration plots are very useful for depth correlation on sand-shale sequences
and also for picking transition zones. Generally these plots are constructed on semi-log
paper with rate plotted in minutes per foot on the horizontal logarithmic scale and depth
on the vertical scale (Figure 4A.32).
A decreasing trend on the minute per foot scale might indicate a change in drillability. In
shale or sand shale sequences this is only possible if the internal (pore) pressure and
porosity of the shale is increasing relative to normal conditions, or possibly a sand is
being drilled. Rate of penetration plots have proven to correlate very well with well logs
and calculated "dc" exponents. Quantification of formation pressure is nearly impossible,
but qualifying the fact that drillability has changed and locating transition zones is quite
easily done with these plots. Rate of penetration plots will be discussed in detail later.
ITT (Interval Transit Time) The ITT is actually a pseudo-sonic log which is generated by
analyzing an ISO-velocity seismic plot. The use of seismic data to predict pore pressure
was first proposed by E. S. Pennebaker, Jr., in his SPE paper entitled, "An Engineering
Interpretation of Seismic Data," published in 1968. This technique was a major
breakthrough in drilling technology, not only because it was a very useful exploration tool,
but because it allowed the well planning engineer to view the subsurface without ever
drilling a single well.
Seismic data is collected by recording echoes from explosive charges, a thump, or an air
gun fired at varying distances from a listening point. This process is repeated many
times with many different listening points. Subsurface horizons cause sound waves to be
reflected and are recorded at the surface as velocity anomalies. Knowing the two way
travel time of the sound wave, it is possible to calculate an interval velocity and therefore,
the interval travel time (T) in micro-seconds per foot. Thus a plot of depth versus T can
be generated. This is a pseudo-sonic log. The process is actually considerably more
complex than this and large computers are necessary to process all of the subsurface
data as it is generated. See Figure 4A.33, for an illustration of an ITT.
A natural problem does exist in the science of interpreting seismic data. Sound waves
traveling through the subsurface tend to echo and re-echo causing multiples (echoes that
reoccur at regular intervals). However, careful examination of an ITT can indicate a
possible transition zone and quantity pore pressure. Seismic processing and specifically
ITT’s are very useful for the drilling engineer and every effort should be made to obtain
this information. Transition zone recognition is only one of several bits of information
which may be obtained. Known formation pressures in a previously drilled well can be
correlated across relatively long distances using several seismic sections. This may give
at least one positive control point for pore pressure in what might otherwise be a
completely unknown environment. Seismic data processing is a fairly complex science
and every drilling engineer should make an effort to obtain as much information on the
subject as possible.
In summary, well planning actually requires an exhaustive research effort on the part of
the drilling engineer. All possible sources of data and information must be employed. It
is simply not always sufficient to drill a new well just as we've drilled the last several
wells. Even with all pertinent information available and the best engineering tools
employed, any well plan is still only a guide and the drilling fluid schedule is only an
estimate. The man drilling the well must use these bits of information as tools and
modify the procedure as the well dictates. Recognizing "real time" indicators of abnormal
pressure and combining these with a highly researched and engineered drilling program
is the key to safe, efficient drilling operations.
Figure 4A.34
Gas cut fluid can, and often does indicate abnormal formation pressure. It is not,
however, always necessary to weight-up the fluid system when an increase in back
ground gas is recorded. Several circumstances need to be considered before any drastic
measures are taken. The abnormal formation pressure may in fact be present but may
not be a problem. Many times tight shale segments may contain gas under pressure, but
because the shale is tight (little or no permeability), the gas will not f low but is drilled up
when the bit penetrates the rock. This does cause an increase in background gas, but
certainly does not constitute a well control problem. It can, and will, if a permeable sand,
under the same pressure considerations, is penetrated by the bit. Circulating “bottoms
up" and observing a return to normal background gas is the general procedure for
handling this type of concern. Other concerns are trip gas and connection gas. In both
cases, an influx of formation gas is noted due to a reduction in bottom hole pressure.
This is caused by the absence of circulating pressure losses when the pump is shut
down, or the swabbing action created when the bit is pulled off bottom.
9. Shale Problems
Shale instability is often caused by an insufficient drilling fluid density. If the internal
(pore) pressure of the shale is not at least balanced by the hydrostatic pressure of the
drilling fluid column, and the shale structure is weak or brittle, it will "pop" into the
wellbore. These relatively large, angular and many times concave pieces of shale will be
very apparent on the shale shaker and can be indicative of abnormal or increasing pore
pressure. This situation may warrant increasing the fluid density or indicate that drilling
should be stopped in order to set casing.
Correct diagnosis of shale instability problems is complicated by the fact that cuttings
nearly identical to those described above can result from poor annular rheology and
hydraulics which cause mechanical erosion. Excessive annular pressure losses
combined with a relatively long open hole exposure time can also cause severe shale
problems. Under these conditions, increasing fluid density will actually compound the
problem. It should be obvious that an accurate assessment of this problem is necessary
prior to making any major changes. Sound preventative measures rather than corrective
measures are really the keys.
It has been well established that as pore pressure increases, without a corresponding
increase in drilling fluid density, the drilling rate will also increase. This is due, in part, to
the fact that abnormally pressured formations are more porous and therefore, less dense
than normally pressured formations.
When all drilling parameters are being held constant and a marked increase in
penetration rate occurs, a drilling break has been experienced. This may happen rapidly
and be very apparent or it can occur gradually. Nevertheless, drilling breaks are most
often the first indicator that a transition from normal to abnormal pressure has occurred.
A well researched drilling program will provide information that will indicate the
approximate depth of the transition zone and make recognition much easier.
The "d" exponent (Figure 4A.35) concept was developed as an attempt to quantify
formation drillability. A simplified drilling rate equation was modified so that an exponent
describing the effect of weight on the bit, and conversely penetration rate, could be used
to indicate a normal shale compaction rate (Figure 4A.36). This then could be used to
locate transition zones and in some cases quantify pore pressure. It has some
shortcomings in that drillability is also affected by hydraulics and mud, bit type and wear,
and formation type (Figure 4A.37). The following equation was used to develop the "d"
exponent.
•
d
WOB
"Normalizes" Changes
( rpm )
Figure 4A.37
PR = k × × e
of WOB and Rotary D
Speed DRILLABILITY (K)
AFFECTED BY:
KW d N e
R =
D
This basic drilling rate equation was modified based on the assumption that it would be
used only in relatively homogeneous shale formations. With this assumption the
formation drillability 'K' was set equal to 1 and the rotary speed exponent 'e' was set
equal to 1 (Figure4A.38). These two assumptions are reasonable, provided that the
formation is homogeneous (shale) and that rate of penetration is directly responsive to
revolutions per minute. In other words, each bit revolution will penetrate one increment
of formation. The resulting equation is:
W d
R = N
D
“d" now is a representative quantifier for formation hardness or drillability. Solving the
simplified drilling rate equation for "d" will yield the desired result.
W d
R = N
D
R Wd
=
N D
Log R Log W
= d
N D
Log R
d = N
Log W
D
Unit conversion constants are inserted so the “d" exponent becomes dimensionless and
the logarithmic expressions are inverted for mathematical convenience. The resulting
equation is:
Figure 4A. 39
Figure 4A.40
Log 60 N
R Gn
d = dc = d
Log 1000 D M.W.
12 W
where:
Gn = Normal formation pressure gradient (expressed in ppg)
M.W. = Actual drilling fluid density (ppg)
One final correction is made which is difficult to justify mathematically, but does account
for effects on drillability caused by drilling fluid properties. Drilling fluid density is
assumed to have the greatest affect on drillability. The calculated “d” exponent is,
therefore, multiplied by the ratio of the normal pressure gradient (usually expressed in
ppg) to the actual drilling fluid density (also expressed in ppg). This is called the
corrected “d” exponent and termed “dc” (Figure 4A.40).
This is a linear correction applied to an exponential function, however, for its intended
use it turns out to be a very applicable tool.
Qualitatively, the "dc" will respond to normal compaction in the same way that resistivity
does. The "dc" exponent will tend to increase with depth through normally pressured
sediments and decrease in under-compacted or abnormally pressured zones. In some
cases, when "dc" data is to be correlated with conductivity or sonic log data, the
reciprocal of “'dc" is multiplied by 100. This generates a"100/dc" plot. This plot of
"l00/dc" will indicate a decreasing trend line through normal pressure, and an increasing
trend in abnormal pressure.
The "dc" or “l00/dc" plot will do an excellent job of identifying a transition zone. It will
however, tend to over-estimate pore pressure as the actual drilling fluid density
increases. This makes the prediction of pore pressure somewhat inaccurate especially in
high pressure environments. Nevertheless, the "dc" exponent is still one of the best "real
time" monitoring tools for changes in drillability and, therefore,. transition zone
recognition.
Generally "dc" exponents should be calculated every ten feet, averaged over each fifty
foot interval and then plotted. Data points will exhibit less scatter if the shale is relatively
clean and homogeneous. Relatively constant weight on bit, rotary speed and hydraulics
will all contribute to a more accurate and reliable plot as they all affect drill rate (Figure
4A.41).
Figure 4A.41
Temperature anomalies.
Factors Effecting Drill Rate Temperature gradient
increases have already been
mentioned and discussed in
• WOB the section on shale property
• RPM trends. At this point, it is
• Hydraulics and Mud sufficient to say that flow line
• Bit Type and Wear temperature will definitely
increase when an abnormally
• Formation pressured environment is
• Differential Pressure drilled. This is due to the fact
that the high pressure
environment is more porous and therefore, acts as a poorer heat exchanger than the
more compacted surrounding sediments.
Heat is actually passed through the more porous sediments much slower and therefore,
creates a higher wellbore temperature when those sediments are penetrated. The
surface response to this phenomena is not immediate, however the information is useful
and flow line temperature should always be monitored.
After drilling the well, every effort should be made to obtain accurate pressure data for
future drilling information (DST data, RFT's, Pressure bombs, Wireline Logs, etc.) (Figure
4A.42).
Figure 4A.42
After Drilling
Gf = Gp + (Go - Gp) Ki
Where:
Gf = fracture gradient (psi/ft, psi, or ppg)
Gp = formation pressure (psi/ft, psi, or ppg)
Go = overburden gradient (psi/ft, psi, or ppg)
Ki = matrix stress coefficient (dimensionless)
Figure 4A.43
Figure 4A.46
V
Ki =
1 - v
It is well to understand the difference between vertical and horizontal fractures and also
between true breakdown (fracture) pressure and fracture extension pressure. A brief
description of each of these follows:
A horizontal fracture is possible at shallow depths and in very hard formations. The
deeper the burial, the harder the formation must be in order to create a horizontal
fracture. When the formation is competent enough to withstand pretrostatic pressure,
fluid entering the formation may lift it vertically, thus creating a horizontal fracture
extending laterally around the wellbore. This type of fracture is extremely rare in drilling,
but has occurred.
Vertical fractures are the most common. Rock will generally fail along a plane which is
perpendicular to the plane of greatest stress. For most depositional environments the
horizontal stresses are greater than the vertical stress, therefore, the rock will have a
tendency to fracture in a vertical plane. The most likely place for any wellbore to fracture
is immediately below the last casing seat. This is based on the fact that if normal
compaction has taken place, formations become harder and more dense as the depth of
burial increases. Therefore, the weakest point will be at the casing shoe. This is an
idealization, and, of course, is not always true.
An important point to consider is that lost returns do in fact occur in shales, not in sand.
This is true because shales are generally weaker than sands. Also, the minimal
permeability in shales will not allow fluid to enter them without causing a fracture. Left to
set, both vertical and horizontal fractures tend to heal themselves in a "soft rock"
environment. However, the time required for the healing process can be quite long.
As wells are drilled, and the time of open hole exposure increases, fluid from the wellbore
gradually seeps into sands and to a lesser degree shales. This seepage increases the
hoop stress around the wellbore and also increases the pore pressure in the near
wellbore area. Understanding these facts certainly indicates that the fracture gradient will
correspondingly increase as well. It is not uncommon to test a casing seat at one leak-
off pressure and later retest it at a higher pressure. (Note that the leak off pressure is
not the same as the fracture (breakdown) pressure, but is still a measure of the
formations strength).
Because of this, the true fracture (breakdown) pressure is generally higher than the
fracture extension pressure (Figure 4A.47). Some test of formation integrity should
always be made (Figure 4A.48). If it is not desirable to go to a formation leak off, a
pressure test of some predetermined magnitude should be performed (Figure4A. 49). In
any event, formation integrity should be estimated prior to drilling, and measured for
verification (Figure4A.50).
There are two recommended methods to predict formation integrity (fracture gradient)
prior to drilling the well and measuring it. The first method of predicting fracture gradients
is from charts developed by Mathews and Kelly or by Eaton. (Figure 4A.51).
Figure 4A.48
• Pressure Test
Figure 4A.50
• Measure
Mathews and Kelly's charts assume a constant over-burden gradient of 19.23 ppg (1
psi/ft) and empirically derived curves for a variable matrix stress coefficient, Ki. The
values obtained were used in the following equation:
FP = PP - Ki ( ob - PP )
Where:
FP = Fracture Pressure (psi)
PP = Pore Pressure (psi)
Ki = Horizontal to vertical stress ratio (dimensionless)
ob = Overburden stress (psi)
Figure 4A.51
Basic Differences
Eaton’s work utilized Poisson’s Ratio to determine the relationship between horizontal
and vertical rock matrix stresses and also used a variable overburden stress gradient.
His work resulted in the following equation:
G ob × D sed
F = PP + V × - PP
1 - v
Where:
FP = Fracture Pressure (psi)
PP = Pore Pressure (psi)
V = Poisson's Ratio (dimensionless)
Gob = Overburden Gradient (psi/ft)
Dsed = Sediment Depth (ft)
This formula has been used to generate the chart shown in Figure 4A.52 and is probably
the most widely used predictive method in the industry today. These methods are similar
in many ways but it is imperative that the use of either of these methods be on a
consistent basis. (i.e. DO NOT attempt to combine the two methods when predicting the
fracture gradient for a proposed well. Doing so can result in large errors.)
Both Mathews and Kelly's method and Eaton's method rely on regionally averaged data
and subsequently are in error for a given specific location. A more accurate technique
involves the calculation of actual overburden stress values from open hole density logs.
The logs may be offset well logs or logs derived from a specific drilling location. This
technique will result in a much more precise fracture pressure prediction for planning
purposes as well as real time prediction while the well is being drilled.
Experience has shown that as we begin to drill in deeper and deeper water, fracture
gradients begin to decrease due to the reduction in overburden pressure (Figure 4A.53).
As we move into deep water, a significant amount of the overburden becomes sea water
rather than soil (Figure 4A.54). This results in a significant loss of available fracture
pressure and can become quite serious when drilling in very deep water (Figure 55).
DTC Technical Memorandum 8802 proposes two methods to predict fracture gradients
when drilling in water deeper than 350'.
Several final summary comments should be made concerning fracture gradient concepts.
Given enough formation information, the overburden gradient, the matrix stress
coefficient and the pore pressure, one can calculate the fracture gradient using the
included equations. Charts are available which graphically represent the same concept.
Also, a relative measure of formation strength can be determined by performing a "leak-
off" test. If done properly, a "leak-off" test does not fracture or break down the formation,
but will indicate that pressure at which the formation will begin to take fluid. "Leak-off"
tests should always be performed with high pressure, low volume pumping units.
Figure 4A.56
Ro 1.2
Gp = Go - ( Go - Gn )
Rn
(Eaton - Resistivity)
Cn 1.2
Gp = Go - ( Go - Gn )
Co
(Eaton - Conductivity)
dco 1.2
Gp = Go - ( Go - Gn )
dcn
(Eaton - d exponents)
Tn 3
Gp = Go - ( Go - Gn )
To
(Eaton - Sonic)
Di - De
Gp = Gn - ( Go - Gn )
(Equivalent Depth)
Di
G f = G p - ( G o - G p ) Ki (Fracture Gradient)
Figure 4A.57
W d
R = K Ne (Fundamental Drilling Equation)
D
W d
R = N (Simplified Drilling Equation)
D
12W d
R = 60N
1000D
60N
d = Log (d exponent Equation)
R
1000D
Log
12W
Gn
dc = d (dc exponent Equation)
Ga
One final idea should be mentioned. The purpose of all well planning when centered
around development drilling projects, is to drill wells safer and more efficient than
previous efforts have allowed. The well plan or design is only a tool to be used as a
guide by the drilling representative on location. True drilling efficiency and optimization
can only occur when the man drilling the well has a strong fundamental knowledge of
pore pressure, and has at his disposal a well researched and engineered drilling plan.
The fundamental equations employed in this research appear in Figures 4A.56 and
4A.57.
*Eaton
**Mathews & Kelly
1. INTRODUCTION
The determination of formation pore pressures from log derived properties is a highly
used and accepted practice in the Gulf of Mexico. Such determinations in other parts of
the U. S. and the world have generally been very difficult, if not impossible, in many
instances. Failure to do so, in most cases, has led to extreme drilling difficulties or
unsuccessful wells. Much of the time, high pressure shale sections have been
misinterpreted as chemically sensitive formations requiring exotic mud chemistries and
resulting in needless excessive expense.
We have developed a technique for the determination of formation pore pressures from
sonic log trends which is universally applicable. This approach has been utilized in
numerous locations around the world with great success. The process will be
demonstrated in detail and several examples of results this achieved from wells around
the world will be presented.
In conjunction with this pore pressure determination process, a simple means of creating
a pore pressure overlay to interpret the data will be demonstrated. This aids in speed of
determinations and simplifies the analysis somewhat.
2. BACKGROUND
Porosity at a given depth is related to the overburden load above. The higher the
overburden, the lower the porosity. At the same given depth and overburden, if
abnormally pressured, the porosity would be higher than for normally pressured rock.
For the same pore pressure increase to be seen at this depth in a lower overburden
environment, we would see a greater porosity increase with respect to a normally
pressured rock accompanying it. Thus, the overburden load directly affects formation
porosity. This in turn affects the relative spacing between pore pressure trend lines in an
overlay. Since the overburden varies from place to place, the trend line spacing varies
with ft. This leads to the need for area specific pore pressure overlays.
The trend line spacings can be developed through determination of the pore pressure
exponents in the Eaton Equation 1. Developing an overlay simplifies the pore pressure
determination process. Here lies the need for the determination of the pressure equation
exponents and the development of an overlay specific for each area.
Pore pressure overlays for any parameter plotted can be simply developed via the Eaton
equations at one known pressure point, preferably two, in a well in the area. The
overburden gradient should also be known, and this can be determined through
integration of a bulk density log from a well in the area. The Eaton equations need to be
rearranged into two formats for overlay development. In Figure 4B.1, we have
rearranged the Eaton pressure equation for plotting resistivity.
Below, the equation is displayed in solving for the pressure exponent (x), and observed
values of resistivity (Ro). Rearranging the Eaton pressure equation for plotting interval
transit time appears in Figure 4B.2.
Investigating the effect the pressure exponent has on the trend line spacing for resistivity
results in Figure 4B.3. Note, that for a given pore pressure, the lower the exponent, the
greater the spacing.
The same analysis for interval transit time appears in Figure 4B.4. Again, the lower the
exponent, the greater the spacing.
The process of creating an overlay first requires solving for the pressure exponent. This
is done by plotting the log data of an offset well in which we have a known abnormal
pressure point. From this we can determine a normal trend line for the parameter
plotted. This normal trend line is extrapolated to the depth of the known abnormal
pressure point to determine the normal value of this parameter. We have the observed
value of the parameter associated with this abnormal pressure point from the log. The
overburden gradient is determined through integration of the density log. We have
everything but the exponent and this is obtained through the equation.
The remainder of the overlay creation process appears in Figure 4B.5. At a given depth,
we assume the pore pressure to be abnormal values in 1 ppg increments and solve for
the observed value of the parameter of interest. These observed values associated with
the respective increments of pore pressure are plotted and trend lines are drawn through
them parallel to the normal trend line established. Thus, we have created a pore
pressure overlay.
The first step is in determining lithology tops. This is done by displaying the gamma ray
and sonic logs in a one inch equals one thousand foot scale. In compressing data like
this, a smoothing function need be applied to avoid a blur of data. Lithology tops are
then determined by picking the points where either the gamma ray or sonic shows a
change in the general trend. This process is illustrated in Figures 4B.6 and 4B.7, with
lithological tops indicated with the dark horizontal lines. The wells utilized in these two
figures are in Indonesia and Norway respectively.
We will illustrate the process on the Indonesian well. The gamma ray and sonic are then
displayed in a one inch equals one hundred foot scale. Again smoothing may be
required. The lithology tops previously determined are translated to this display. Sonic
velocity trend lines are then drawn on the sonic log with respect to the shale readings
within lithological sections as illustrated in Figure 4B.8.
The sonic velocity trend lines are then drawn on semi-logarithmic paper honoring
lithology tops as in Figure 4B.9.
These lithology tops become recalibration points in this process. In this, the sonic
velocity trend in one lithological section is traced. The velocity trend is recalibrated by
shifting the tracing over at the lithology change, joining the last value of interval velocity in
the last lithological section with the first value in the next. This results in a continuous
relative interval velocity profile as in Figure 4B.10.
For this well we have a known formation pore pressure at 3800 feet of 10.6 ppg
equivalent mud weight. We integrate the bulk density log and determine the overburden
gradient. We now have what we need to solve for the pore pressure exponent and
create an overlay for the area. This has been done as previously described in the
creation of pore pressure overlays in Figure 4B.11.
Determination of pore pressure for all formations can now be read directly from the
overlay as in Figure 4B.12. Note that in the intervals which appear to have been drilled
under balanced, extreme difficulties with shale sloughing were encountered. All the
formations encountered lacked permeability, except at TD where the mud weight had to
be raised to exceed the pore pressure.
In the California well illustrated on the previous page, we have two known abnormal
pressure points. At 4900 feet we have an 11.2 ppg and at 5900 feet we have a 12.4 ppg
pore pressure in mud weight equivalents. We solve for the exponent at 4900 feet where
we have the 11.2 ppg. Using the exponent derived, we solve for pore pressure at 5900
feet where we know the answer. As can be seen in Figure 4B.13 we get a pore pressure
of 12.34 using the exponent in the Eaton pressure equation. Thus, we have confidence
in the establishment of our normal trend line for this well and the determination of the
pore pressure exponent for this area.
Again we can now create an overlay for use in all wells in the general area. By assuming
values of abnormal pressure in increments of one ppg we solve for observed values of
sonic velocity utilizing the rearranged Eaton equation for sonic velocities as in Figure
4B.14.
The overlay can be applied to the recalibrated sonic velocity trend to determine pore
pressures for all formations as in Figure 4B.1 5. Note we had an additional known
pressure point of 8.5 ppg at 9050 feet which falls appropriately on the overlay.
Important Note: It is important to recognize which side of the sonic log trend lines need
be drawn. In many instances, it may be necessary to change from one side to the other
upon crossing lithology tops. In Figure 4B.16, we have picked lithology tops for a well in
the Gulf of Mexico.
We find in this well, from close examination of the sonic response with respect to the
shales, that between the depths of 9300 and 10,400 feet, it becomes necessary to switch
from plotting trend lines on the right to the left side of the sonic as illustrated in Figure 4B.
17.
In this Gulf Coast well, if we use the regionally averaged pore pressure exponent of 3 in
the pore pressure equation, we determine pore pressures to be a bit high as in Figure
4B.18.
If we determine an area specific exponent and overlay for this well as described, we
obtain better accuracy as illustrated in Figure 4B.19.
5. EXAMPLES OF RESULTS
The technique has been used on numerous wells around the world with great success.
Some examples appear in the figures to follow. Figure 4B.20 represents the results for a
Gulf of Mexico well in 2300 feet of water.
The results for another well offshore Norway appears in Figure 4B.24.
Information Required:
1. Sonic logs from all Offset Wells displayed in True Vertical Depth with Gamma Ray
and/or SP, preferably both. One set of logs should be displayed with a depth
scale of 1" = 1000' and a second set displayed with a depth scale of 1” = 100'.
Note that a smoothing function will be required for the 1” = 1000' log. Scale for the
sonic should be linear with the majority of the plot using all of the area available
(i.e. all four tracks).
2. Information as to where logging changes and casing points occurred (i.e. depths).
3. Any and all geologic data (i.e. location of faults in the area, cross-sections,
structure maps, etc.).
4. Mud, Bit and Drilling records from all offset wells to be analyzed.
Procedure:
1. Using the 1” = 1000 ft logs, identity intervals on the log where an abrupt shift in
sonic and/or gamma ray indicate a lithology change that is effecting the sonic.
Draw a horizontal line through these points. These lines are referred to as "re-
calibration points”.
2. Transfer the points identified in #1 to the 1” = 100 ft logs and note any changes in
log runs and casing points. Examine the log further to identify any additional
lithology changes that might have been missed in #1. Draw horizontal lines
through these points.
3. Draw trend lines connecting the sonic response in shales between the re-
calibration lines drawn in #2. Note: it is very important that the trend lines are
drawn on the correct side of the sonic plot. Examine the gamma ray plot and the
sonic plot to determine the sonic response to the shales as compared to the other
lithology types. If the sonic response for a shale is to the left of that for the
surrounding non-shale lithologies, the trend lines should be drawn on the left side
of the plot. Conversely, if the shale response is to the right then the trend line
should be drawn on the right side of the plot. The determination as to which side
of the plot the trend lines are drawn should be made for each interval between the
recalibration points as the relationship between shales and non-shales can change
with depth. The reason for the need to chose the correct side of the plot is that the
slope of the trend line can change from one side of the plot to the other. If the
wrong side is used it can result in errors for the remainder of the plot and a
decrease in accuracy for the pore pressure prediction.
4. Transfer the trend lines and re-calibration points onto two cycle semi-log paper.
The easiest way to do this is to note the sonic values at the top and bottom of
each trend line along with any inflections between the re-calibration points. Then
plot the same values on the semi-log paper. The typical depth scale used on the
semi-log paper is 1” = 1000 ft.
5. Overlay the semi-log paper with a second sheet of semi-log paper. Trace the
trendlines adjusting the top piece of paper to account for the shifting required to
connect the trendlines across the re-calibration points. It is very important that the
two pieces of paper maintain the same orientation when shifting!!
6. Examine any additional data available in order to determine where the normal
trend line should be placed. Extend the normal trend line to the bottom of the plot.
7. If a regional overlay or an overlay from another offset wall is being used, place the
overlay on the plot created in #6, aligning the respective normal trend lines. Skip
to step #9.
8. If an overlay is to be developed, note the depth and known pressure point(s) onto
the plot. Use Eaton's equations to determine the exponent and then calculate the
observed values for the various pore pressures. Plot these values onto the semi-
log paper at the same depth for which the exponent was determined. Draw the
trend lines through these points parallel to the normal trend line determined in #6.
9. Read the pore pressure values at the inflection points and plot them on a pressure
vs. depth plot on regular coordinate paper.
1. INTRODUCTION
Techniques for estimating formation pore pressures from relative changes in log derived
shale properties have been in the industry and accepted for years. The basic premise, of
course, is that, at depth, shale porosity is a function of the pore fluid pressure and the log
derived shale properties are a function of the shale porosity. There are difficulties,
however, in that there are other factors which can influence the log properties of shales
than just porosity alone. For this reason, the determination of formation pore pressures
from log properties can be difficult and inaccurate.
Recalibration techniques and considerations have been developed for the various log
derived properties typically used in the industry, however, this paper will focus on those
affecting shale resistivity, or conductivity if you prefer. In particular, the effects of
changes in formation salinities and the effects of multiple log runs are the significant
factors which have been addressed and dealt with. As increasing pore pressures are
encountered, shale resistivities decrease as an indication of increasing pressure.
However, if the formation water salinity increases, the shale resistivity will also drop,
complicating the analysis. Also, as we set each string of casing, we log each
subsequently smaller hole with a different log tool, probably a different logging engineer
and in many cases with an entirely different logging unit. These changes which occur at
each log run also result in difficulty in pore pressure determinations across these
intervals.
This method addresses the effect of changing salinities and log runs and enables one to
accurately determine formation pore pressures from shale resistivity and conductivity
trends. The technique has been utilized on hundreds of wells throughout the Gulf of
Mexico with great success. Typically pore pressure determinations are within a few
tenths of a pound per gallon of measured pore pressures in the adjacent virgin sands.
2. BACKGROUND
A method of determining formation pore pressure by analyzing log property trends was
developed by Ben Eaton. With this technique we can plot shale values on a semi-log
scale and determine a normal trend line through these values in the normally pressured,
normally compacted section of the hole, as in Figure 4C.1.
By comparing shale properties which deviate from this normal trend to the values, at
depth, of the normal trend line, an estimation of pore pressure is made. The calculation
of formation pore pressure is made by utilizing the equations derived by Ben Eaton and
these equations appear in Figure 4C.2.
A more simplified approach results if a pore pressure overlay is created based on the
equation to be utilized. Such an overlay for Gulf Coast shale resistivities appears in
Figure 4C.3.
With use of the overlay, the normal trend line of the overlay (9 ppg for the Gulf Coast) is
aligned through the normally pressured normally compacted section of the hole and as
shale properties deviate from this normal trend, a value of pore pressure is more simply
read from the overlay as in Figure 4C.4.
The Problem
If the only factor affecting the shale resistivity were the porosity of the shale, then the
pore pressure determinations would be quite simple. However, we know that the
formation water salinity becomes a factor in that changes in salinity cause changes in
shale resistivity. The effects on shale resistivity that salinity can cause is illustrated in
Figure 4C.5.
In addition, as we log each section of hole we use a different tool, a different logging
engineer, etc., and abrupt changes in shale resistivity can result as illustrated in Figure
4C.6. We must, therefore, compensate for these factors if we are to accurately
determine formation pore pressures from log properties.
When we analyze a typical selection of shale resistivity values as plotted we find we get
an erratic selection of points as plotted in Figure 4C.7. Note that the determination of the
position of the normal trend line can be quite difficult and may be very subjective. We
have to make a representative trend to determine a normal trend line.
Another approach to the selection of the data points appears to be to smooth the data on
the log itself. This is done by drawing trend lines on the resistivity curve itself as in
Figure 4C.8. The resistivity trend lines are drawn with respect to the shale resistivities
only. The shale values which are plotted are taken from the trend lines with all inflection
points on the trend being honored. This allows for a smooth selection of data points
which enables an easier determination of the normal trend as was seen previously in
Figure 4C.4.
The recalibration process of these same data points appears in Figure 4C.10. The
overlay is applied to the data aligning the 9 ppg trend line through the interval of points
which represent the normally pressured, normally compacted section of the hole. Pore
pressures are then read directly from the overlay below the normally pressured section
down to a recalibration point. Recalibration across this change in log runs is performed
by shifting the overlay horizontally, without allowing any rotation, until the last reading of
pore pressure in the previous section coincides with the first reading of pore pressure in
the next section.
In reference to Figure 4C.10, the last pore pressure reading was 12 ppg at approximately
8200 feet. In this illustration, the overlay is shifted until the 12 ppg trend line is over the
first valid shale resistivity reading. The recalibration process has taken place and values
of pore pressure for this next section of hole are again read directly from the overlay to
the next recalibration point.
Changes in formation water salinity also require the same recalibration procedure. We
have a log section in Figure 4C.11 where such an abrupt salinity change occurs. At
approximately 9520 feet, we cross a fault and see a significant salinity change. Note the
dramatic change in shale resistivity above and below this point. As we plot the resistivity
values as in Figure 4C.12, we indicate this point of salinity change as a recalibration
point.
The recalibration process is the same as before, shifting the overlay at this point as in
Figure 4C.13.
In Figure 4C.5 we saw that in areas of low formation water salinity, a small change in
salinity has a significant influence on the shale resistivity readings. In these
environments, it is recommended that sonic logs be relied on for pore pressure analysis
unless a detailed formation salinity analysis is per-formed from surface to total depth.
However, in areas of higher formation water salinities, such as the Gulf Coast, small
changes do not have a dramatic impact on shale resistivity values, but large variations
do.
With this in mind, the primary concern is in recognizing when a significant salinity change
has occurred. To recognize such an occurrence, a simple technique can be applied. SP
trend lines can be drawn as well as resistivity trend lines on the log as in Figure 4C.14.
The SP baseline will generally be affected significantly by two things; changes in salinity,
and changes in shale porosity due to variations in pore pressure. When the formation
pore pressures are increasing, the SP baseline will drift to the right in conjunction with
resistivity drifting to the left. When this occurs the logging engineer will adjust the SP
baseline back to the left in order to prevent the SP curve from entering the depth track.
Below this logger's adjustment, the SP will continue its right hand drift as pressure
continues to build. This phenomena is illustrated in Figure 4C.15.
Therefore, the recommended practice is to first draw trend lines on the SP curve, noting
shifts in this trend for possible recalibration locations, then to repeat the process of
drawing trend lines on the shale resistivities. After plotting shale resistivity values,
overlay recalibration is performed only where the SP base line has shifted due to
formation salinity changes as illustrated in Figure 4C.17.
If an overlay is not being used, these same recalibrations can be performed at changes
in log runs and salinity by shifting the normal trend line an appropriate amount. This is
process is illustrated in Figure 4C.18.
These techniques have been utilized on hundreds of wells throughout the Gulf of Mexico
with tremendous success and excellent accuracy. Typical results thus achieved are
illustrated in Figures 4C.19 and 4C.20.
Information Required:
1. Resistivity Logs for all Offset Wells displayed in True Vertical Depth with Gamma
Ray and/or SP, preferably both. Scale should be 1” = 100 ft for the depth track
although 5” = 100 ft will work. Scale for the resistivity should be linear with the
majority of the resistivity plot using all of the area available (i.e., all four tracks).
2. Information as to where logging changes and casing points occurred (i.e., depths).
3. Any and all geologic data (i.e., location of faults in the area, cross-sections, structure
maps, etc.)
4. Mud, Bit and Drilling records from all offset wells to be analyzed.
Procedure:
1. Note all log run changes and casing points on the resistivity logs. Examine the logs
to determine any additional points where shifting is necessary (i.e., SP shifts).
2. Determine the shale response trend lines between the shift points. Trend lines for
resistivity logs should always be drawn along the left side of the log trace.
3. Determine the resistivity values at each shift and inflection point and transpose the
data onto two cycle semi-log paper noting the shift points (i.e., recalibration points).
4. Overlay the semi-log paper with a second piece of semi-log paper of the same scale.
Trace the trend lines adjusting the second piece of semi-log paper to account for the
shifting required (i.e., connect the trend lines together by sliding the top sheet of
paper), it is very important that the two pieces of paper remain the same orientation
when the shifting is performed!!
5. Examine the drilling data available for the well in an effort to accurately determine
where the normal trend line occurs. Pick the normal trend line and extend it to the
bottom of the plot.
6. If the regional overlay or an overlay from another offset well is being used, place the
overlay on the plot created in #5 aligning the normal trend lines. Go to step #8.
7. If an overlay is to be created from the plot, note the depth and known pressure
point(s) onto the plot. Use Eaton’s equations to determine the exponent and then
calculate the observed values for the various pore pressures. Plot these values onto
the semi-log paper at the same depth for which the exponent was determined. Draw
trend lines through these points parallel to the normal trend line determined by #5.
8. Read the pore pressure values at the inflection points and plot them on a Pressure
vs. Depth plot on regular coordinate paper.
1. INTRODUCTION
The approach will be illustrated in detail, and the results of its use on a number of wells
will be presented, and explained. Mud weights required to drill these wells will also be
illustrated as well as any pressure control data such as a kick.
2. BACKGROUND
In just about any environment, there tends to be a mixture comprising a rock's makeup.
Generally speaking, in all sands there is some shale, in all shales there is some sand, in
all carbonates there is some sand and shale and so forth. Thus, when we speak of a
rock as being a sandstone for instance, we are referring to its primary compositional
nature.
When we analyze a gamma ray log in a classical sand shale environment, we look for a
gamma ray response to the right for shale indication, and to the left for an indication of
sandstone.
When we analyze the gamma ray in a carbonate environment, the peak responses to the
right are considered to be the result of shale within the rock matrix. It will be in these
intervals where we will concentrate our efforts.
3. THE TECHNIQUE
The first step is in determining lithology tops. This is done by displaying the gamma ray
and sonic logs in a one inch equals one thousand foot scale. In compressing data like
this, a smoothing function needs to be applied to avoid blurring the data. Lithology tops
are then determined by picking the points where either the gamma ray or sonic shows a
change in the general trend.
This process is illustrated in Figure 4D.1, with lithological tops indicated with the dark
horizontal lines. The well utilized in this Figure is in the Destin Dome area of the Gulf of
Mexico.
The gamma ray and sonic are then displayed in a one inch equals one hundred foot
scale. Again smoothing may be required. The lithology tops previously determined are
translated to this display. Next to this data we display an unsmoothed version of the
gamma ray, as well as an SP, resistivity and conductivity curves as in Figures 4D.2
through 4D.6.
Within lithological sections, we analyze the gamma ray peaks which trend to the right in
the shale direction. We've circled these in Figures 4D.2 through 4D.6.
With respect to these gamma ray peaks to the right, we draw sonic log velocity trend
lines, honoring the sonic velocities which correspond to the gamma ray intervals
previously circled.
These corresponding sonic velocities have also been circled in Figures 4D.2 through
4D.6 and the corresponding trend lines have been drawn.
Note that in some instances in Figures 4D.2 through 4D.6 that the velocity trend lines
appear on the left of the sonic log, and in others on the right.
The sonic velocity trend lines are then drawn on semi-logarithmic paper honoring
lithology tops as in Figure 4D.7.
These lithology tops become recalibration points in this process. In this, the sonic
velocity trend in one lithological section is traced. The velocity trend is recalibrated by
shifting the tracing over at the lithology change, joining the last value of interval velocity in
the last lithological section with the first value in the next. This results in a continuous
relative interval velocity profile as in Figure 4D. 8.
Note, in Figure 4D.9, that we can draw a normal trend line through the normally
pressured, normally compacted section of the hole from 5400 to 8700 feet. For this well
we have a known formation pore pressure at 21,000 feet of 15.2 ppg equivalent mud
weight.
We integrate the bulk density log and determine the overburden gradient. We now have
what we need to solve for the pore pressure exponent and create an overlay for the area.
This has been done in Figure 4D.10.
Determination of pore pressure for all formations can now be read directly from the
overlay. This has been done with results graphically displayed in Figure 4D.11. Note
that in the intervals which appear to have been drilled under balanced, difficulties with
torque and drag were encountered.
All the formations encountered lacked permeability, other than the formation at 21,000
feet which was of low permeability. This formation of approximately 15.2 ppg pore
pressure continued to flow at a rate of roughly 3/4 a barrel per hour with a mud weight as
high as 14.9 ppg in the hole. The bottom portion of the well experienced a pressure
regression and mud weights could be reduced.
It is important to recognize which side of the sonic log trend lines need be drawn. In
many instances, it may be necessary to change from one side to the other upon crossing
lithology tops. In Figures 2 through 6, from close examination of the sonic response with
respect to the gamma ray intervals selected, it becomes necessary to switch from
plotting trend lines on the right to the left side of the sonic and vice-versa.
4. EXAMPLES OF RESULTS
The technique has been used on numerous wells with great success. Figures 4D.12
through 4D.17 depict some of the results.
For an oil field to develop, there are several things nature must have provided. There
must be a source of the hydrocarbons. Hopefully these hydrocarbons migrate through a
reservoir type rock, having permeability and porosity, to a point where they become
trapped in some way by a capping formation.
Some typical traps seen around the world are in anticlinal structures, against
impermeable formations such as salt domes, or against impermeable formations across
a fault.
We gain an understanding of the look of the formations below the earth’s surface through
the use of seismic surveys. An energy source is used to pulse sound waves through the
earth. Some of the energy of these sound waves is bounced back at each formation top
or lithological change as in Figure 4E.1.
The results of this seismic survey yields a picture which approximates the generalized
shape of the formations below the earth’s surface as in Figure 4E.2.
The composition of formation clays can be significantly dependent upon the depth of
burial as in Figure 4E.3. As clays are subjected to ever increasing pressures and
temperatures due to deeper depths of burial resulting from additional sedimentation, they
experience a metamorphosis. Montmorillonites convert to illites at deeper depths and so
forth. The well on the right can penetrate a shale formation with the clay composition as
shown. Upon drilling an offset well on the left, if this same stratigraphy is encountered at
some shallower depth, the shales may have undergone less change due to a lower
overburden load and temperature and we might see a somewhat different clay
composition. This can have an effect on the drilling fluid chemistry required to drill this
formation.
Sands which have developed as the result of an ancient river channel are characterized
by a ratty low permeable development in the upper portion, with a gradually improving
development and permeability toward the bottom as in Figure 4E.4.
Sands which have developed as the result of an ancient beach are characterized as tine
grained on bottom and coarse on top. These sands tend to gradually thin out and lose
development as we head seaward to the right in the illustration in Figure 4E.5.
Many, if not most, ancient carbonates were deposited simultaneously in three different
macro-environments; shelf, slope and basin as illustrated in Figure 4E.6. The shelf
environment consists of broad, shallow seas, mostly less than 100 feet of water.
Currents are weak, so generally lime mud has been deposited. Usually there is enough
current or wave motion to keep the water oxygenated. Scattered isolated coral heads or
larger patch reefs are common. Sometimes in mud banks, oxygen is used up and
organic matter is preserved.
The material on the slope consists of lime sands and blocks that have been broken off
the reef by waves and deposited in strata with an initial dip. They are called reef talus
and sometimes form excellent reservoirs.
The material in the basin is fine grained, usually lime mud. Normally, it does not have
sufficient permeability to produce hydrocarbons. In a few places, chalk has accumulated,
formed from the tiny shells of algae called coccoliths. They have considerable porosity
but very low permeability. The basinal carbonates often grade laterally into shale. In the
case of epi-continental basins, it often happens that there is little circulation of the water
in the deeper parts of the basins. Organic matter is preserved because not enough
oxygen is brought in to destroy it. Occasionally such deposits become highly organic and
may become source rocks of hydrocarbons.
If fractured formations are encountered such that the fracture are in tension due to
folding, then these fractures may be open and a source of taking fluids as in Figure 4E.7.
However, if these same formations are in compression at another location, the fractures
may be closed.
Many times, due to the stresses within the earth, a shearing of the rocks occurs as in
Figure 4E.9. This results in formations on either side of the fault being at different subsea
depths and formation fluids on either side not being in hydraulic communication.
Figure 4E.11 is a seismic section which also illustrates the same information as the
previous illustration in the bed thickening. Seismic lines tying proposed wells to any
offsets provide the drilling man with tremendous information in anticipating formation tops
and relative changes in stratigraphy between wells.
Figure 4E.12 illustrates some important definition of terms. When discussing a well plan
with the geologist and geophysicist, it is important for the drilling engineer to define
terminology. Many times the same wording has two different meanings to different work
groups. In the strictest definition, formation pressure is equivalent to pore pressure.
When the drilling man speaks of pore pressure, he refers to it in values converted to mud
weight equivalents.
Normal pore pressure, generally speaking, is equivalent to the native fluid gradient in the
area. Strictly speaking, abnormal pressure is anything but normal, however, the drilling
man refers to anything above normal as abnormal, anything below as subnormal.
Because we speak of pressures in mud weight equivalents, lower bottom hole pressure
does not equate to lower pore pressures, and for this reason, drilling the same
stratigraphy as an offset does not equate to drilling the same pore pressures. ITT curves
are tools generated from surface seismic data. They are essentially synthetic sonic logs
used to predict pore pressures.
Figure 4E.13 illustrates the effect dipping beds have on pore pressure in normally
pressured environments, in that there is no change.
Figure 4E.14 illustrates the effect hydrocarbons have on increasing pore pressure as we
move up dip. Note that the bottom hole pressure is lower but the pore pressure is higher
in mud weight equivalent up dip.
The higher the pore pressure environment, the greater the increase in pore pressure
seen as we move up dip as illustrated in Figure 4E.16.
The deeper the environment, the less severe the changes which are seen as in Figure
4E. 1 7.
Figure 4E.18 summarizes the effects dipping beds have on pore pressures to be seen.
As we move from well to well and cross a fault, Figure 4E.19, there can be a resulting
change in pore pressure. Post depositional faults occur after a significant portion of the
formations involved have been deposited. If the sealing mechanism forms very rapidly,
we can see higher pore pressures on the down thrown side of the fault due to the
increased amount of overburden supported above the seal. Note, however, in this
illustration, that the faulting example utilizes a fault of 2000 feet of throw. Be aware that
a fault of lesser throw would result in a proportionally lower change in pore pressures
seen. Consequently, relatively small faults generally do not impact pore pressures
significantly.
If the sealing mechanism forms very slowly with respect to geological time, as in Figure
4E.20, which is the more likely case, the opposite may be true in that the up-thrown side
is the higher pore pressure. This results from the greater overburden load on the down-
thrown side causing greater fluid volumes to be squeezed up and out through this slowly
developing seal.
Depositional faults are those developing gradually along with the deposition which is
taking place as in Figure 4E.21. It the sealing mechanism forms very rapidly, with respect
to geological time, a similar result is seen in that the down thrown side of the fault can be
the higher pore pressure.
Again with depositional faults, as with post depositional, if the seal formed gradually, the
up thrown side can be the higher pore pressure as well, the more likely case, as in Figure
4E.22.
Reverse faulting can have dramatic effects on pore pressures as in Figure 4E.23.
Pressures are trapped at deeper depths and brought shallower. this results in
significantly higher pore pressures in the up thrown block in many instances.
Figure 4E.24 is a general summary of the impact on pore pressure that faulting can have.
Figure 4E.25 is a display of a gamma ray and sonic log. Logs contain valuable
information in the way of pore pressures for the drilling man. Changes in log property
trends are indicative of the top of abnormal pressure as well as the magnitudes thereof
below. Note the dramatic change in the trend of the sonic log at 12,800 indicating the top
of abnormal pressure. Note that above this point there is a definite trend of the sonic
velocities. This trend could be projected below to compare to the abnormal sonic
velocities. A comparison of the normal trend and the observed values in the abnormally
pressured environment can be equated to a pore pressure value.
Illustrated in Figure 4E.26 are the equations used to compare the normal trends of a log
property in the normally pressured, normally compacted section of the hole, extrapolated
down, to the observed readings in the abnormally pressured section of the hole.
From the pore pressure equations, a pore pressure overlay can be created to determine
pore pressures. Figure 4E.27 is such an overlay for resistivity.
A pore pressure overlay is applied to the data by aligning the normal trend line of the
overlay through the data points of the normally pressured normally compacted section of
the hole as in Figure 4E.28. Values of abnormal pressure are then read directly from the
overlay in the sections below.
Pore pressures are determined in offset wells and projected to a proposed location for
well planning. In the seismic illustration of Figure 4E.29 we evaluate the pore pressures
seen in the down dip well in the center of the seismic trace. We are concerned with what
to expect in the updip wellbores proposed to the left. We will look at the updip directional
well to illustrate how this works.
If we merely designed this up dip well to utilize the casing program which was seen to be
adequate in the down dip well, then the casing program proposed would be that on the
left of Figure 4E.30. However, due to the encountering of all stratigraphies significantly
up dip, we would find ourselves stopping short of the same stratigraphic horizon achieved
in the offset and would actual see the scenario on the left occur. Let’s examine why?
Figure 4E.31 is a graphical illustration of the pore pressures and fracture gradients
determined to exist in this offset down dip well. Note the top of abnormal pressure is
seen at approximately 8800 feet. With surface casing set at 4003 feet, the maximum
mud weight utilized handles the pore pressures seen in this well without exceeding the
surface casing shoe fracture gradient.
Many times we have differential pressure limitations which dictate a maximum mud
weight which can be safely used in a hole section. In Figure 4E.33, here too in the
proposed well, differential limitations would dictate the need for an additional string of
pipe.
The effect dipping beds has on the magnitude of pore pressures is further complicated by
the resulting effect on formation integrity, or fracture gradient. Figure 4E.34 is a typical
fracture gradient chart. To illustrate, if we look at a normally pressured rock at 6000 feet,
we would expect a fracture gradient of 15.3 ppg. At this same depth, if the pore pressure
were 10 ppg, the fracture gradient expected would be 15.6 ppg. In other words, for a 1
ppg increase in pore pressure at this depth, we only gain.3 ppg increase in fracture
gradient. We will demonstrate the effect this has in the following illustrations.
Consider the case of an existing offset well and a proposed well located slightly up dip.
Figure 4E.35 is a graphical display of the pore pressures and fracture gradients
determined to exist in the off set well. Note the top of abnormal pressure is at
approximately 5500 feet.
As we determine the depths at which we will encounter the horizons in the proposed well
and estimate pore pressures to be seen, we find the top of pressure will be up dip at
approximately 4800 feet as in Figure 4E.36.
Illustrating the impact on fracture gradient, we see that for significant increases in pore
pressures, small increases in expected fracture gradient result as in Figure 4E.37. This
results in a narrower band of limits between which we have to select casing points.
Once a prediction or pore pressures and fracture gradients is made, we can then
determine, as far as pressures are concerned, casing points. For instance, in Figure
4E.38 we can see that a minimum of 3200 feet of surface casing is required in order to
reach a depth of 5500 feet.
In Figure 4E.39, we see that if we chose to set surface casing at an arbitrary depth of
3370 feet, then two strings of intermediate casing would be required to reach a depth of
10,000 feet.,
In Figure 4E.40, we see that if surface casing were set at a depth of 4745 feet, then one
intermediate string of casing would be required to reach a depth of 10,000 feet.
1. INTRODUCTION
An accurate prediction of formation pore pressures prior to drilling a well can significantly
improve drilling performance, reduce the number of costly drilling problems encountered,
and in many instances, can be the determining factor in successfully reaching objectives.
This is particularly true when dealing with exploratory wells. The fact that we are going to
penetrate the same stratigraphies as offset wells does not equate to being in the same
pore pressure environment. In fact, when dealing with exploratory wells, there is a high
percentage of the time when this is not the case.
This approach to predicting formation pore pressures utilizes offset log information, and
velocity information from surface seismic data. All surface seismic data has velocity
information built into ft in the way of stacking velocities. From these stacking velocities,
estimates of interval velocities can be made. We have developed a quantitative
technique for determining formation pore pressures from the relative changes in these
velocities.
From the stacking velocity data, we develop curves we refer to as ITT (Interval Transit
Time) curves. ITT curves are generated at offset locations as well as the proposed
drilling location. Actual offset well pore pressures are determined from the logs and the
ITT curves at these offsets are calibrated to the known pressures. After this calibration
process, the ITT at the proposed location can be interpreted to predict formation pore
pressures to be encountered.
The combination of these tools yields a reliable pore pressure scenario from which to
plan a drilling program and execute the operations. This procedure has been utilized on
numerous wells throughout the Gulf of Mexico as well as a number of wells around the
world. The technique will be illustrated with Figures 4F.1 through 4F.14 as a step by step
analysis is done. Results of the use of this methodology will also be illustrated.
2. BACKGROUND
A Mr. Pennebaker did some research back in the sixties on interval velocities vs. depth
for various geological ages in normally pressured environments. The results of his work
indicate trends which appear to be straight lines on a log-log scale (see Figure 4F.1
below).
If we take one of these curves, such as that for the Pliocene age, and re-plot the data on
a semi-logarithmic scale, with the vertical linear scale for depth in feet, and the horizontal
log scale for interval velocity in microseconds per foot, we obtain a curve like the one in
Figure 4F.2. This indicates a curved relationship with respect to depth for interval velocity
in a normally pressured, homogeneous environment.
Procedure
ITT curves are generated at the offset wells as in Figure 4F.3. Lithology tops are
indicated with respect to the ITT curve based on where they are seen on the gamma ray
sonic log. Keep in mind that slight differences in the depths of these lithologies may exist
on the ITT vs. the gamma ray sonic, since the velocities in the ITT are not accurate
interval velocities. A comparison of the ITT to the gamma ray sonic log is necessary to
select these lithology tops on the ITT. Some of the less significant velocity shifts across
lithology tops will be ignored, but the more dramatic ones will be honored.
Velocity trend lines are then drawn on the ITT curve within lithological sections. It is
important to stay on the same velocity side of the ITT curve as was determined to be
necessary in the pore pressure determination process while plotting the sonic log of the
offset well. This has been done in Figure 4F.4. Note that only the significant velocity
shifts have been honored.
The ITT velocity trend lines are traced within lithological sections. Upon arriving at a
significant velocity shift at a lithology top a recalibration is performed. This is done by
shifting the trace and aligning the last velocity value of the trend in the previous
lithological section, with the first velocity value of the next. Tracing of the velocity trend
line in the next lithological section is then done until a continuous velocity trend profile is
obtained. Recalibration of the ITT velocity trends appears in Figure 4F.5.
Pore pressures are then determined from the ITT and calibrated to the known pore
pressures in this offset well. This is done with several possible interpretation
approaches, two of which will be illustrated here. In the first interpretative approach, we
take a typical Pennebaker curve such as the one in Figure 4F.2, and align it against the
ITT curve to identify the normally pressured, normally compacted interval as in Figure
4F.6. The ITT curve departs from this Pennebaker curve at approximately 5300 feet.
We then take the recalibrated velocity trend line curve of Figure 4F.5, and draw a straight
normal trend line through the velocity trend down to this depth of 5300 feet. We then
create a pore pressure overlay for use on ITT curves in the area. We solve for the
pressure exponent using a known pressure point in this offset well. As in Figure 4F.7, we
then, at the same depth of this known pore pressure point, assume pore pressures in
increments of 1 ppg, and solve for observed values of interval transit time. Through
these observed values of interval transit time, parallel trend lines are drawn to the normal
trend line previously selected.
A comparison of pore pressures determined from the ITT and the mud weights used on
this offset well appear in Figure 4F.8.
Note, between the depths of 5300 and 8800, that the ITT is reacting to the porosity seen
in this interval, Figure 4F.9, and not to pressure. Lithological information such as this
needs always to be kept in mind during the interpretation process.
An alternate approach is illustrated in Figure 4F.10. Here, a normal velocity trend line is
drawn through the apparent normally pressured, normally compacted interval
immediately above the top of pressure as determined from plotting the sonic log on this
well. The same procedure of solving for the pressure exponent, and creating an overlay
for the ITT as before is followed.
A comparison of pore pressures from the ITT utilizing this approach, and the mud
weights used to drill this well appear in Figure 4F.11.
To insure a quality ITT pore pressure overlay has been created, and proper solution of
the pressure exponent, the overlay should be used on the ITT of another offset well. In
Figure 4F.12, we have an ITT on another offset well with velocity trend lines selected.
Application of the overlay, developed from the second approach, to the recalibrated
velocity trend lines appears in Figure 4F.13.
A comparison of the pore pressures from the ITT on this well to those determined from
the sonic log, and the mud weights required in drilling appear in Figure 4F.14. Since we
have good agreement, we feel confident in the overlay creation process, and believe it to
be good for ITT curves, generated with the same logic and rationale in the area.
An ITT is generated at the proposed location and lithological tops which dictate a
recalibration in velocity trends are correlated through seismic ties and indicated as in
Figure 4F.15.
The velocity trends are recalibrated into a continuous curve as in Figure 4F.17.
The ITT is interpreted at the proposed location using the same reasoning which was
successful at the offsets. In the first approach, as in Figure 4F.18, a Pennebaker curve
is aligned with the ITT curve at the proposed well. identifying the normally pressured
normally compacted section.
Again a straight normal trend line is drawn through the velocity trend in this interval. The
ITT pore pressure overlay created in this approach is used on the recalibrated velocity
trends as in Figure 4F.19.
Pore pressures for the proposed well are then read directly from the overlay, the results
of which appear in Figure 4F.20.
In the second approach, again, a normal trend line is drawn through the stratigraphic
interval immediately above the top of pressure, based on correlation with the offsets
through the seismic ties. The overlay created with this method is applied to the
recalibrated ITT velocity trend lines as in Figure 4F.22.
Again pore pressures are read directly from the overlay, the results of which appear in
Figure 4F.23.
REFERENCES
1. INTRODUCTION
M I S S I N G
1. INTRODUCTION
All surface seismic data has velocity information built into it in the way of stacking
velocities. From these stacking velocities, estimates of interval velocities can be
determined. We have developed a quantitative technique for determining formation pore
pressures from the relative changes in these velocities.
From the stacking velocity data, we develop curves which we refer to as ITT (Interval
Transit Time) curves. The ITT curves are generated at the proposed location and are
correlated to offset location curves to calibrate velocities. Offset location pore pressures
are predicted and are combined with structural geology to predict anticipated pore
pressures of the proposed wellbore. With these curves, in conjunction with geological
information to determine structural changes, and offset logs to determine pore pressures
in offsets, we can predict pore pressures to be encountered in a proposed wellbore.
Significant dollars can be saved and increased success in reaching objectives can be
realized when pore pressures can be anticipated prior to drilling a well. This technique
offers the opportunity to do so.
We have developed techniques whereby surface seismic data can be utilized to aid in the
prediction of formation pore pressures prior to the drilling of a well. Although these
techniques have been developed for many drilling environments, this technique focuses
on the sand shale depositional environment.
2. BACKGROUND INFORMATION
Since seismic information has velocity information built into the data by the very nature of
how it is recorded, in the process of stacking the data, stacking velocities are determined.
From these velocities, interval velocities can be determined. Since all seismic
information contains this velocity information, all geophysical companies have the
capability to generate a curve of interval velocity vs. depth at any point on a seismic line.
This, essentially, becomes a synthetic sonic log. Mr. Pennebaker (1) did some research
back in the sixties on interval velocities vs. depth for various geological ages in normally
pressured environments. The results of his work appears in Figure 4H.1.
Since the trends appear to be straight lines on this log-log scale, we merely extrapolated
the data to deeper depths as in Figure 4H.2. If we take one of these curves, such as that
for the Pliocene age, and re-plot the data on a semi-logarithmic scale, with the vertical
linear scale for depth in feet and the horizontal log scale for interval velocity in micro
seconds per foot, we obtain a curve like the one in Figure 4H.3, where we observe a
curved relationship with respect to depth for interval velocity in a normally pressured
environment.
If we generate a curve at a given shot point on a seismic line of interval velocity with
respect to depth, we get something like that in Figure 4H.4. We've generated this curve
on a semi-log scale, the vertical being linear for depth in 1” = 1000', and the horizontal
scale is logarithmic for interval velocity in microseconds per foot. This curve is generated
utilizing both stacking velocities and seismic amplitudes, and this is something all
geophysical companies can produce. After the curve has been generated, smooth 'trend
lines" are physically drawn onto the curves representing the compactions in the velocity
trend for lithologies encountered. This occurs at a point where a significant lithological
change takes place.
At this point, in the well, we start to see grain to grain contact in the formations. As we
look further down the curve we find there are no additional abrupt shifts in the curve
indicative of additional significant lithological changes. Looking at another such curve in
this type of environment in Figure4H.6, and again drawing trend lines on the curve as in
Figure 4H.7, we find similar results.
We will refer to these interval velocity curves as Interval Transit Time (ITT) curves.
We've generated an ITT curve at a location where we knew the top of abnormal pressure
to be fairly deep, Figure 4H.8, in order to compare this ITT in a normally pressured
environment to one of Pennebaker's normal velocity profiles such as the one in Figure
4H.3. If we compare the velocity trend from Pennebaker in Figure 4H.3 to the curve in
Figure 4H.8, but display the Pennebaker information to the left a bit so the two curves do
not overlap, we arrive at Figure 4H.9.
We have done several other things with Figure 4H.9 as well. We have divided the depth
intervals into different lithological sections with the horizontal dashed curves, and,
through these separate lithological sections, we have drawn trend lines with respect to
the interval velocity curve. We note that these trends of the interval velocity in each
lithological section are parallel to the Pennebaker curve. We also note that up shallow in
this sand shale environment, that these trend lines may have an abrupt shift either to the
left or right as we change lithologies.
However, as we go deeper there are no lithological changes which cause these abrupt
shifts in interval velocity. From many such observations at various locations in sand
shale environments, we have concluded that the general shape of the interval velocity
curve is what is curve. We use stacking velocities only, and run a smoothing function on
the data which yields a very smooth ITT curve.
Figures 4H.10 and 4H.11 represent comparisons of “smoothed" ITT curves against
normal ITT curves. This is a necessary comparison to be sure the data has not been
poorly manipulated as is illustrated in Figure 4H.12 where the smooth version is in poor
agreement with the original.
3. PROCEDURE
First, we select offset locations with quality log information. At these locations, we
generate an ITT curve as in Figure4H.16. Here the very smooth curve is the ITT
generated at a shotpoint on a seismic line which is as close to the well bore as possible.
The other curve, with many wiggles, is a sonic log from the same well. The ITT is
calibrated to the sonic as closely as possible since the sonic actually measures interval
velocities in this offset wellbore, and the ITT approximates them from the surface seismic
data. All the data is displayed on two cycle semi-logarithmic paper.
The vertical depth scale is at one inch equals one thousand feet and the horizontal log
scale is in microseconds per foot. Next we generate pore pressure plots on the logs from
this off set and determine actual pore pressures for as many intervals as possible. We
display this pore pressure data next to the ITT curve as in Figure 4H.17.
Next, we will construct a straight normal trend line through the ITT curve and calculate
pore pressures using Eaton's equation. We will realign this normal trend line to whatever
position is necessary until the pore pressures derived from the ITT match those derived
from the logs. This is done through trial and error and may take several attempts. In
Figure4H.18 we have our first attempt at drawing the normal velocity trend. We find with
this normal trend, that the pore pressures we calculate from the ITT are too high.
We know from Eaton's equations, that as we bring the observed and normal values of a
parameter closer together, that calculated values of pore pressures will get lower. With
this in mind, we draw a second normal trend line parallel to the first but shifted to the right
as in Figure 4H.19. Now we find that the top of abnormal pressure coincides, ITT
determined pore pressures at bottom are accurate, however, ITT pore pressures through
the mid-section of the well are still too high.
If we rotate this normal trend line of Figure 4H.19 about the point of the top of abnormal
pressure as in Figure 4H.20, we find ITT indicated pore pressures are more accurate
through the midsection of the well but are too low toward the bottom. Keep in mind that
the interval velocity data determined from stacking velocity information is not precise and
consequently the ITT may be off depth and may require a depth adjustment.
In Figure 4H.21 we go back and analyze the comparison of the ITT to the sonic log of
this offset well. We see at approximately 10,200’ that the sonic breaks to the left while
the ITT drifts right. Shortly below, the two curves drift in opposite directions. To analyze
the required depth adjustment, a light table would be required where the two curves
could be handled separately allowing independent movement for proper alignment. In
Figure 4H.22 we have the sonic and as we adjust the ITT with respect to it, we finally find
we get a proper match in Figure 4H.23 by shifting the ITT 1700' upward.
Now the two curves break to the same directions at the same depths. We do not
normally find we require this much adjustment in most cases, but this dramatic example
points out the need for care. Now that we know the depth adjustment requirements we
proceed with another attempt at drawing the normal velocity trend as in Figure 4H.24,
realizing that pore pressure values calculated from the ITT require a 1700' adjustment.
We find that two more attempts are required as in Figures 4H.25 and 4H.26 before we
finally arrive at a good match. We now find that pore pressures calculated from the ITT
very closely match log derived pore pressures. We now have a calibration on the normal
velocity trend through this stratigraphic interval for this particular drilling area. To be sure
this is the case, we will graphically compare the results.
In Figure 4H.27 we have a plot of pore pressures vs. depth from both the log properties,
and from the ITT with no depth adjustment to the ITT values. As we depth adjust the ITT
values, we arrive at Figure 4H.28. Graphically, we see a close match and feel
comfortable with the calibration of the normal velocity trend.
An alternate approach to the calibration of the normal velocity trend line can be more
simply performed by backing into, so to speak, the Eaton pressure equation. In Figure
4H.29, we have an ITT curve for an off set well for which we know pore pressures from
our log analysis. These known pore pressures are listed on the left in the Figure at
depth.
Since we know the pore pressures, the value of the pressure exponent, and the
observed values of interval travel time ( the ITT values ), we can rearrange the equation
and solve for normal values of interval transit time. These would be the values of interval
transit time for normally pressured, normally compacted rock of the same type at the
same depth. Once solving for the normal values, we can then curve fit the data points
with the best straight line fit as in Figure 4H.30.
Now we generate an ITT curve at the proposed location as in Figure 4H.30. We use the
same normal velocity trend as calibrated in the offset for the stratigraphic interval of
interest. We use the Eaton equation for interval transit time again to compare values of
observed to normal interval transit times and calculate pore pressures as in Figure
4H.31. We realize these values require a depth adjustment as did the off set and plot
them as such in Figure 4H.32.
To illustrate the value of such prediction techniques, we have included some results. In
Figures 4H.34 through 4H.39, we graphically compare predictions of pore pressure
utilizing this ITT technology, to both actual mud weight requirements, and actual log
derived pore pressures as the wells were drilled. In each case you can see that
predicted pore pressures very closely matched well requirements, and these are only a
few examples of the excellent results we've seen.
4. NOMENCLATURE
Gf = Fracture gradient
v = Poisson's ratio
5. REFERENCES
Information Required:
2. ITT curves for the offset wells and the proposed location. Note, If the offset and/or
proposed wells are highly deviated, more than one ITT curve may be required at
these locations.
3. Sonic logs ( 1” = 100 ft) interpreted for pore pressure at the offset locations.
Procedure:
1. For the offset wells, invert the ITT curve and place It onto the 1” = 1000’ sonic log.
Correlate the ITT curve with the sonic log and transfer the re-calibration points from
the sonic log onto the ITT. For the proposed location, correlate the re-calibration
points form the nearest offset, if possible. If correlation is not possible then pick re-
calibration points based on abrupt changes in the ITT curve.
2. Draw the trend lines for the “shale" responses between the re-calibration points.
Note: You will need to examine the interpreted sonic log (1” = 100’) to determine
which side of the ITT curve should be plotted. Since the ITT is typically displayed
opposite of the sonic, if the sonic trend line was drawn on the left for a particular
interval. then the trend lines for that interval will be drawn on the right on the ITT.
The ITT curve will not have a gamma ray displayed. On offset wells you can lay the
ITT next to the 1" = 100' sonic log and read the shale zones from the gamma ray,
This will require correcting the ITT curve for depth, respective to the sonic log. For
the proposed location, you will need to assume that the ITT curve response is the
shale response. Note: the ITT curve for the proposed well will typically have the
same share response (i.e. left or right side of the curve) as the offset ITT curves.
3. Transfer the trend lines from the ITT curve onto two cyclo-semi-log paper. If the ITT
curve is displayed on a semi-log scale identical to your semi-log paper, you can
overlay the semi-log paper onto the ITT curve and trace the trend lines directly (a
light table works great). The re-calibration points should also be transferred. Note
that it is advisable to record the ITT depth and the sonic depth determined from the
correlation since the ITT is not depth accurate.
4. Overlay the semi-log paper with a second sheet of semi-log paper and trace the
trend lines adjusting the overlaid sheet to account for the shifting required to connect
the trend lines across the re-calibration points. It is very important that the two
pieces of paper maintain the same orientation during the shifting!!
5. Correct the ITT plot for depth by making an adjust identical to that required for
the ITT at the offset well (i.e. if the ITT was corrected up 1000’, at the offset,
correct the ITT 1000’ up at the proposed location.)
6. Determine the normal trend line for the ITT plot. The top of pressure for an offset
well should be the same as that for the sonic plot. Remember to correct the ITT
plot for depth!!
7. If an overlay is to be created, note the depth and known pressure point(s) onto the
plot (remember to plot these pressure point(s) on the sonic depth, not the ITT
depth). Determine the exponent using Eaton's equations and then calculate the
observed values for various pore pressures and plot these values on the semi-log
paper. Draw trend lines through these values parallel to the normal trend line
determined in #5.
8. Read the pore pressure values at the inflection points from the overlay and plot
them on a pressure vs. depth plot on regular coordinate paper. Compare the
results of the ITT analysis for the proposed well, that of the pore pressures
determined from correlating the sonic pore pressures across the seismic line. If
the ITT curve compares favorably to the correlation, use the correlation. The
correlation is used because it typically has more "characters”.
Note: It the ITT and the sonic plots have a similar profile but the depths differ slightly,
adjust the sonic plot to match the ITT. The reasoning behind this is that the ITT has
already been adjusted for depth (step #5). We assume that the correlation is in error due
to fluid densities different than the “native” fluid density that was assumed when the
correlation was made.
If the ITT and the correlation are not similar, use the ITT. This is an indication that there
is not a hydraulic relationship between the wells. In some instances the ITT and the
correlation are similar for a portion of the well but deviate at some point. In this instance,
use the correlation for that part of the well where the predictions are similar and use the
ITT for the remainder of the well.
1. INTRODUCTION
This approach to prediction of formation pore pressures incorporates into it the geological
model. Offset log information, structural relationships between wells from surface to total
depth, and velocity data from surface seismic, are utilized to model the anticipated pore
pressures to be seen in the wellbore from surface to total depth. This modeling approach
yields a reliable pore pressure scenario from which to plan a drilling program and execute
operations.
Numerous wells throughout the Gulf of Mexico as well as a number around the world
have been analyzed with this approach with great success. The modeling approach will
be illustrated in this writing as well as results of the use of this methodology.
2. BACKGROUND
When dealing with abnormally pressured environments there are some definite structural
impacts on the pressures to be seen in a proposed well with respect to the offsets,
assuming there is a hydraulic relationship between the wells.
greater overburden load on the down-thrown side causing greater fluid volumes to be
squeezed up and out through this slowly developing seal.
The difficulty arises in that sometimes wells which are proposed do have a hydraulic
relationship with respect to the off sets and sometimes they do not. Other times the
relationship exists only for a portion of the well. This approach identifies when and for
what portion of the well the relationship exists and handles the overall predictive process.
The approach will be demonstrated by illustration of the steps taken on three example
wells. In the first example, the proposed well is determined to have a hydraulic
relationship with the offsets. In the second example, it does not. The third example
utilizes a proposed well which appears to have a hydraulic relationship with the offsets
only to a specific formation, below which it does not.
In Figure 4I.1, we have a surface base map which shows the location of a proposed well
and the offsets.
We obtain interpreted seismic lines from the geophysicist which tie the proposed well to
these offsets, as in Figures 4I.2 and 4I.3.
For every formation in the offset wells for which a pore pressure is determined, a
correlation established with the proposed well as in Figures 4I.6 and 4I.7.
At this point, we assume there is a hydraulic relationship between wells. Using the native
fluid gradient for the area, we calculate anticipated pore pressures at the proposed
location for all formations correlated. This is done by taking the difference in the depth
below datum the formations are seen in each well and correcting bottom hole pressure
for that formation by the gradient of the native fluid for the area. This has been done in
Figures 4I.8 and 4I.9.
Figure 4I.8
Figure 4I.9
We now have predicted pore pressures for the proposed well assuming a hydraulic
relationship between wells exists as in Figure 4I.10.
We then generate ITT ( interval transit time ) curves at the offset wells as well as at the
proposed location. These ITT curves are generated from the surface seismic data. They
are essentially synthetic sonic logs which approximate interval velocities with respect to
depth. The ITT curves generated at the offset wells are calibrated to the sonic logs as
closely as possible as in Figure 4I. 11.
Once convinced that this calibration is satisfactory, an ITT curve is generated at the
proposed location utilizing the same programming steps and considerations as in Figure
4I.12.
The ITT curves at the offsets are calibrated to the known pore pressures to exist in these
offsets as in Figure 4I.13, until a match between pore pressures derived from the ITT and
those derived from log properties is obtained as closely as possible as in Figure 4I.14.
We then interpret the ITT curve at the proposed location using the same logic or
reasoning which was successful at the offsets as in Figure 4I.15.
Next we compare pore pressures predicted from the ITT at the proposed location to
those predicted by correlation which assumed a hydraulic relationship between wells to
exist as in Figure 4I.16. Here we see that the two curves match quite nicely which
verifies the hydraulic relationship.
However, the fluid gradient assumed may have been slightly different or correlations may
have deviated slightly yielding the difference seen between the two curves.
Consequently, the correlative scenario is shifted upward to coincide with the ITT scenario
as in Figure 4I.17, and this newly placed correlative scenario becomes the predicted pore
pressure curve for the proposed well.
Fracture gradients are then determined and casing points selected as in Figure 4I.18.
The actual mud weights required to drill this well as well as actual pore pressures seen
vs. predicted pore pressures appear in Figure 4I.19.
In Figure 4I.20 we have a structure map for an area with a proposed location just
southeast of center.
An available offset well appears to the north east of this location. Pore pressures
determined for this offset well appear in Figure 4I.21.
We correlate all formations for which pore pressures have been determined to the
proposed well through the seismic ties as in Figure 4I.22.
The ITT is interpreted at this offset and calibrated to the known pore pressures in Figure
4I.24, the results of which appear in Figure 4I.25.
To be sure this calibration process has been done satisfactorily, we graphically compare
pore pressures from the ITT to those determined from log properties in Figure 4I.26.
Graphically comparing these results to the correlative scenario yields Figure 4I.28. From
this comparison we see that the two curves disagree entirely indicating that a hydraulic
relationship between wells does not exist.
The ITT predicted pore pressures become the only scenario from which to plan the
proposed well. This is done in Figure 4I.29.
A comparison of pore pressures predicted for this well and those actually seen based on
log properties appears in Figure 4I.30.
A comparison to actual mud weights required during drilling appears in Figure 4I.31.
In Figure 4I.32, we have selected lithology tops for an offset well from the gamma ray
and sonic logs.
Sonic velocity trends within lithological sections are then determined as in Figure 4I.33.
Pore pressures for all formations are then determined for this offset as in Figure 4I.34
and are compared to mud weights used in Figure 4I.35.
Again for as many formation as possible, correlations are established between wells
through seismic ties (not shown for this example). Pore pressures are calculated for the
proposed well assuming a hydraulic relationship as in Figure 4I.36.
Figure 4I.36
continued as 4sec_I_b.doc
Lithology tops are indicated with respect to the ITT curve based on where they are seen
on the gamma ray sonic log. Keep in mind that slight differences in the depths of these
lithologies may exist on the ITT vs. the gamma ray sonic since the velocities in the ITT
are not accurate interval velocities. A comparison of the ITT to the gamma ray sonic log
is necessary to select these on the ITT. Some of the less significant velocity shifts
across lithology tops will be ignored, but the more dramatic ones will be honored as in
Figure 4I.39.
Pore pressures are then determined from the ITT and calibrated to the known pore
pressures in this offset well as in Figure 4I.41.
A comparison of pore pressures determined from the ITT and the mud weights used on
this offset appear in Figure 4I.42.
Note, between the depths of 5300 and 8800, that the ITT is reacting to the porosity seen
in this interval, Figure 4I.43, and not to pressure. Lithological information such as this
need always be kept in mind during the interpretation process.
An ITT is generated at the proposed location and lithological tops which dictate a
recalibration in velocity trends are correlated through seismic ties and indicated as in
Figure 4I.44.
ITT velocity trends are determined as before in Figure 4I.45 and are recalibrated into a
continuous curve as in Figure 4I.46.
The ITT is interpreted at the proposed location using the same reasoning which was
successful at the offsets as in Figure 4I.47, and the results of which appear in Figure
4I.48.
We then compare this ITT prediction to the correlative scenario as in Figure 4I.50. Here
we see that down to a depth of approximately 13,500 feet, the two scenarios agree, and
below they do not. This suggests a hydraulic relationship between wells to 13,500 feet
exists and below they are in separate pressure environments.
We, therefore, use the correlative information down to the depth of 13,500 feet and the
ITT prediction below yielding Figure 4I.51 as our predicted pore pressure profile for the
well.
Adding fracture gradients, we arrive at Figure 4I.52 for the purpose of planning this well.
Some basic rock mechanics definitions appear in Figure 4J.1. These are the general
terms used in formation integrity analysis. In determination of formation integrity, it is
often necessary to analyze the principal stresses within the rock and the shearing stress
for failure. The definition of these terms, therefore, appears in Figure 4J.2. A summary
of some of the ideas of Hubbert and Willis with respect to the principal stresses in
sedimentary rocks appears in Figure 4J.3. Some of the more pertinent equations used
by Mathews and Kelly and Eaton in formation integrity analysis appear in Figure 4J.4,
and the Eaton pore pressure equations and fracture gradient equation which we
commonly use appear in Figure 4J.5.
The matrix stress coefficient, being a function of Poisson’s ratio, varies with depth and
the overburden gradient. Poisson’s ratio for the shallow waters of the Gulf Coast area
appears in Figure 4J.9 The matrix stress coefficients for the same shallow waters of the
Gulf Coast based on a regionally averaged variable overburden gradient appear in Figure
4J.10, as per Ben Eaton. The matrix stress coefficients for the shallow waters of the Gulf
Coast based on a constant overburden gradient of 1 psi per foot as per Mathews and
Kelly appears in Figure 4J.11. The regionally averaged variable overburden gradient for
the shallow waters of the Gulf Coast as per Ben Eaton appear in Figure 4J.12.
When we use Eaton’s fracture gradient equation, for a given depth, regardless of
formation pore pressure, we plug in a value of matrix stress coefficient for that depth.
Theoretically, therefore, the matrix stress coefficient is predominantly dependent on the
overburden gradient. With this in mind, knowing the overburden gradient we might be
able to estimate a pseudo-matrix stress coefficient. If we know the overburden gradient,
the matrix stress coefficient, and an accurate prediction of formation pore pressure, we
can then estimate the formation fracture gradient.
For example, in Figure 4J.13, we have data for a well drilled in 2130 feet of water.
Consider the following, let:
G Fma - G Pma
Kp =
G Oaa - G Pma
where:
Kp = pseudo matrix stress coefficient
G Fma = the normal fracture gradient for the area for normally pressured rock
G Pma = the normal pore pressure gradient for the area
G Oaa = the actual overburden gradient for the area
Figure 4J.13
GFnap - GPna
KP = GF = GP + (G O - GP )K
GOaa - GPna
From the example illustrated, we see from knowing the actual overburden gradient we
can estimate a pseudo matrix stress coefficient. Once estimating a pseudo matrix stress
coefficient, we can utilize it in the Eaton fracture gradient in conjunction with the actual
overburden gradient and estimated pore pressures. Comparing results thus obtained to
actual leak off tests for this well in 2130 feet yields excellent results.
The key to this technique is obviously in knowing the overburden gradient for the area.
Theoretically, an estimate of this is possible from the velocity information built into
surface seismic data. More empirical work in this area needs to be done, however, the
technique appears to be promising.
Information Required:
3. Plot of matrix stress coefficient for the area. This can be obtained
from Eaton’s plot for the Gulf Coast, from a sonic waveform analysis
log run on an offset well, or estimated from leak off test data on other
offset wells.
PROCEDURE
3. Using Eaton’s equation for Fracture gradients, input the values into the
equation and determine the fracture gradient (in psi/ft).
1. INTRODUCTION
The occurrence of stuck pipe can significantly increase the cost of drilling, decrease, if
not eliminate, the success in reaching objectives, and negate attempts to retrieve
valuable formation evaluation data. We have developed many tools and techniques in
the oil field to solve problems such as stuck pipe after they occur, but the real key to
savings and success is to avoid problems.
Approximately 600 wells were analyzed throughout the Gulf Coast. This survey included
trouble-free wells, and those which had experienced differential and mechanical sticking,
as well. From this data, we are able to predict the environments which are likely to have
a high risk of stuck pipe occurrences. Once knowing the risk factors, we are able to
either design the well to avoid the high risk situations or to plan for it accordingly.
The results of this statistical analysis will be presented in a simple format which enables
one to estimate the risk of stuck pipe occurrence for a given well. Use of the information
for both well planning and successful execution will be explained and demonstrated.
Examples of the occurrence of problems when entering high risk areas will also be
illustrated. During the courses of drilling some 100 wells, with attention to the guide lines
to be presented, only eight incidences of stuck pipe were encountered.
2. RESULTS OF ANALYSIS
In Figure 4K.1, we have a curve which estimates the risk of a differential sticking
occurrence. This is for wells drilled as straight holes with water base mud. It assumes
formation permeabilities fall within gulf coast average values for any depth. This curve
suggests that when the differential pressure in the wellbore, the differential between the
mud hydrostatic and the formation pressure, begins to exceed 2000 psi, the probability of
differential stuck pipe goes up extremely rapidly. It suggests that differentials much
higher than 2000 psi should be avoided. Of course, if lower than Gulf Coast average
permeabilities are expected, then this threshold limit would be higher, and vice versa. It
also assumes that mud properties are not out of line. That is, solids problems, poor
rheological characteristics, etc. could result in a lower threshold limit.
We will consider 2000 psi to be the threshold limit for straight holes with water base
muds when drilling virgin pressured formations. Investigation of directional wells
indicates a lower limit exists. Figure 4K.2 illustrates the formula for calculating this
threshold limit for directional wells. The curve would be the same as that of Figure 4K.1,
but shifted to the left to the value of the threshold limit.
Figure 4K.2
The estimate of the risk of occurrence of mechanically stuck pipe requires a bit more
elaboration. Figure 4K.3 illustrates a general guideline. In the Figure, 95% success
maximum safe angle equates to a scenario which has a 5% risk of a mechanical stuck
pipe occurrence. Eighty-five percent success maximum safe angle equates to a scenario
which has a 15% risk of a mechanical stuck pipe occurrence.
If a well is planned with anticipated mud weight of greater than 14 ppg at total depth, will
reach a depth greater than 15,000 feet, and is planned as a 30 degree directional well, it
has a 15% mechanical risk. For wells which are planned with angles greater than the
85% success maximum safe angle, calculation of the additional mechanical risk is
illustrated in Figure 4K.4. The 85% success angle is subtracted from the magnitude of
the desired angle to obtain an additional risk angle. The sine of this risk angle multiplied
by 200 yields the percent of additional mechanical risk to be expected. These
mechanical risks can b4e reduced by a reduction of the length of the open hole sections
by addition of another string of casing to the program. On the surface, this would seem
to be a costly addition, however, statistically greater expenditures are made in dealing
with mechanical stuck pipe situations in long open hold sections.
Figure 4K.3
• These guidelines are for the build up and hold angle type directionals
• S-curves are always considered more risky
Figure 4K.4
Example:
• For a mud wt. Range at TD 14.0 ppg and a range of TD (MD) 15,000’
• The (85% success max. poss. angle) = 30°
• If (Desired angle) = 50°
• Then (50°) - (30°) = 20° = Risk Angle
85% success max. poss. angle = 15% risk, therefore, this proposal
would have 15% risk + 68% additional risk or 83% chance of failure
Note: These risks are based on standard casing programs and could be
reduced through use of additional casing strings at intermediate depths
but, this will greatly increase costs.
Figure 4K.5
These risk factors are for the build and hold angle type of directional wells. S curve type
wells add additional m4echanical risk. We shall define hang down as the length of hole
below the point where the angle begins to drop. The additional risk this drop off presents
is illustrated in Figure 4K.5.
3. HISTORICAL EXAMPLES
In Figure 4K.6, we have a comparison of a proposed casing program vs. That actually
needed. The anticipated mud weight at total depth was __ ppg, proposed total measured
depth was 15,558 feet, and the prescribed directional angle 50 degrees. From Figures
4K.3 and 4K.4, we calculate a mechanical risk factor of 83% (the example used in Figure
4.) During the course of drilling this well, several incidences of mechanically stuck pipe
were encountered necessitating costly fishing operations, and an unanticipated additional
string of intermediate casing in order to reach authorized depth.
In Figure 4K.7, we have the predicted pore pressures and fracture gradients for a well
with predicted required casing points due to differential pressure limitations. Also
displayed are the mud weights used on the well, as dictated by hole conditions. Below
the first intermediate string of casing, the need for a second string of casing due to
differential limitations was tested. Upon exceeding the threshold mud weight (threshold
differential pressure) by two to three tenths of a ppg, the drill string became differentially
stuck on two attempts necessitating two sidetracks and an additional string of
intermediate casing.
As a normal course of action in the well planning process, the formation pore pressures
and fracture gradients are predicted. Baring any other considerations, casing points
would be selected by working down the hole between these two curves as in Figure 4K.8.
In this illustration, after setting surface casing, we drill till we reach a point where our pore
pressures approach within .5 ppg (a safety margin of our anticipated fracture gradient at
the surface casing shoe. We plan to set pipe here and select the next casing point in the
same fashion and so forth. If we analyze further, we examine the anticipated pore
pressures of all permeable formations exposed in each section of hole and calculate the
maximum allowable safe differential pressure we can risk by use of the equation in
Figure 4K.2. This determines the maximum mud weight we can afford to expose
formations to and, in turn, to a casing point selection as in Figures 4K.9 and 4K.10.
Once determining anticipated casing points for a proposed well, the occurrence of
faulting at these locations presents some interesting problems. In Figure 4K.12, we have
a log section. The sand from 13,000 to 13,100 feet is our last normally pressured 9 ppg
formation. The next sand seen at 13,350 feet is of a 12.5 ppg pore pressure. As a
straight hole, 2000 psi is the differential limitation.
This equates to a maximum allowable mud weight of 12 ppg to be used across the sand
at 13,000 feet. Consequently, the shale interval between these two sands is a definite
casing point. In Figure 4K.13, we have a correlative ROP (rate of penetration) plot for an
adjacent well next to the log section of the previous Figure. Note the apparent kick in the
midst of this transitional shale section.
It was desired to set casing 100 feet below the last 9 ppg sand since approximately 250
feet of transitional shale is available as a casing point window based on the offset well
illustrated in the previous Figure. Unfortunately, in this well, a fault of approximately 200
feet of throw was encountered at this crucial casing point as illustrated in Figure 4K.14.
Weighting up for the 12.5 ppg pore pressure sand resulted in differentially stuck pipe.
In Figure 4K.15, we have the anticipated pore pressures illustrated for a well. Note that
at approximately 11,500 feet a fault is anticipated which results in an abrupt increase in
pore pressure. With surface casing setting depth for this well at 4005 feet, anticipated
formation integrity is 14.2 ppg.
Note that upon crossing the fault anticipated, if the first formation seen were to have
permeability, mud weights required to gain well control would exceed the surface casing
shoe integrity, a very dangerous situation. Safety considerations require a protective
intermediate string of casing be set some reasonable distance above the fault to be
encountered as illustrat4ed. Plotting the differential pressures which would be seen in
the well due to mud weight requirements results in Figure 4K.16.
Note that even with the protective intermediate string set, differential pressures exceed
the limit upon crossing the fault if weighting up for permeability is required immediately
upon crossing it. Another intermediate string is required in crossing the fault and we can
only hope the first formation encountered is impermeable to avoid weighting up prior to
setting it. Obviously, from this illustration, faults at casing points can make for very high
risk expensive wells. The required casing program appears in Figure 4K.17 and, as can
be seen, two strings of pipe are required to deal with this fault at a crucial casing point. If
permeability is seen immediately upon crossing it, differentially stuck pipe is still likely to
occur.
Faults at casing points should be avoided via changing surface locations or alteration of
directional plans, whenever possible.
Use of these statistics for well planning is strongly recommended. However, the arbitrary
setting of casing as prescribed is not to be implied. In Figure 4K.18, we have a casing
point selection plot for a well which requires one of its intermediate strings due to
differential limitations.
The casing program appears in Figure 4K.19 and, as can be seen from the illustration,
one string was eliminated during the course of drilling the well.
In Figure 4K.20, we have a log section from an offset to this well with a correlative ROP
for the well in question. Note below the 11-3/4” casing string, that the permeable sands
of concern in the offset well from 12,550 to 12,850 feet are shaled-out in our well based
on the ROP plot. If permeable sands are not seen, differential sticking cannot occur.
Hence, the mud weight limitation no longer applies. We, therefore, drill beyond this
casing point based on correlative information gained during the course of drilling the well.
When drilling wells in high pressure environments, there is a tendency to want to push
each string of casing as deep as possible. Indeed, there are times when this is
advantageous and desirable. However, the pore pressure and fracture gradient
predictions should be studied very closely to determine if this effort is justifiable. Many
times these efforts are merely costly and inefficient. For example, in Figure 4K.21, we
have illustrated a typical pore pressure and fracture gradient profile of a well. As we look
closely at this illustration, it can be seen that an intermediate string of casing is
necessary to reach authorized depth in any event.
With an anticipated surface casing shoe test of 14.6 ppg, and a pore pressure at total
depth of 16 ppg, it is inevitable. The differential limitations of a 9 ppg formation exposed
at 12,000 in a straight hole dictates a maximum allowable mud weight of 12.2 ppg. This
would dictate a casing point at 12,400 feet. Pushing this string deeper into higher
pressure is possible with a surface shoe integrity of 14.6 ppg, however, we risk
differential sticking needlessly, and very little is gained since the fracture gradient
changes little at this intermediate string. We also increase the length and cost of the
intermediate string to be run to no advantage. In this example and in many other cases,
upon close examination, there are only disadvantages to pushing the intermediate casing
strings beyond the point of the differential pressure limitations.
Application of this information is possible in other parts of the world if we keep in mind
the area for which it was developed in mind, and how it relates geologically to these other
area. For instance, the differential sticking statistics are based on average Gulf Coast
permeabilities. If we encounter higher permeabilities than this, lower differential
pressures would become a problem and vice versa. Average permeabilities for the Gulf
Coast appear in Figure 4K.22. Mechanical statistics would have to be related by
comparison of rock composition.
1. INTRODUCTION
In Figure 4L.1, we have an electric log on a well. The last nine pound per gallon pore
pressure sand is seen at 13,000 feet. The next sand seen of a 12.5 pound per gallon
pore pressure at 13,350 feet. The base of the last nine pound sand is at 13,100 feet.
We have approximately 250 feet of shale between these two sands. To drill the 12.5
pound sand would require a mud weight in excess of this. This mud weight against a
nine pound sand at 13,000 feet greatly exceeds the 2,000 psi of differential we know to
be a problem. We cannot drill into this 12.5 pound sand with this higher mud weight
without becoming differentially stuck. If we set pipe above the nine pound sand, we
would still be faced with the same problem of having this nine pound sand exposed when
we drill this 12.5 pound sand.
In Figure 4L.2, we have an ROP plot on an offset well drilled and, approximately 60 to 70
feet below this nine pound sand, we take a kick.
In Figure 4L.3, we have the electric log on the first well on the left and this ROP plot is
spliced next to it on the right. The last nine pound sands are correlated together. Based
on correlation, we are taking a kick in the middle of the so-called shale section or
window.
With logs on both wells side by side, we see in Figure 4L.4 that this shale section, in the
well taking the kick, is no longer 250 feet thick, but rather on the order of 60 feet thick.
This is due to a fault encountered in this well faulting out approximately 200 feet of this
window, or shale section, bringing this 12.5 pound sand closer to this nine pound sand.
$900,000 in expenses were incurred in straightening the situation out on this well.
In Figure 4L.5, we have a structure map with a proposed location, well #120, shown
approximately at center. To the southeast of this location, we see a very massive fault,
of approximately 1000 feet of throw, moving down to the southeast.
We analyze a cross section for this well, Figure 4L.6. The casing window for this
environment is between the D10 and D12 sands. We can see along this cross-section
that we can be drilling wells where after we come out of the D3 sand, when we cross the
fault, the first thing we see could be the D14 sand. The D14 and D16 sands get up in the
16 pound range. If all we have set is surface casing and the first thing we see is
permeable sand, we have serious problems. We never know what we will see first when
we cross a fault. We may be lucky and see some high pressure shale to give us
indicators. However, the first thing we see may be a permeable sand in which case we
have extreme difficulties.
If the pressures are high enough, we can have some serious well control and safety
problems on our hands. One approach to handle this is to move the wellbore over and
drill this well directionally such that we cross this fault as depicted in Figure 4L.6 at a
much shallower depth, thereby allowing us to see the casing window between the D10
and D12 sands. It would be unusual for a drilling man to propose a 40 degree directional
as in Figure 4L.7, but there are times when this may be a better alternative to an
additional string of casing.
This is not to say that all faults at casing points are problems. It is a question of whether
or not the magnitude of the throw of the fault significantly exceed or is very close to the
casing window in which we have to set pipe.
In Figure 4L.8, we have a very busy cross section with several faults moving through the
area. We have a series of sands, D7, D8, D9, D10, and then a D11 series. In the case
of this cross-section, the casing window is between the D9 and D11A sands. We see,
that even though a fault cuts through this section and faults out the D10 sand, that there
is still several hundred feet of shale between the D9 and D11A sands.
Therefore, a fault at this casing window is not a problem since the throw of the fault is
significantly less than the window in which we have to set pipe. Faults at casing points
are a problem only if they exceed the window in which we need to set pipe.
In Figure 4L.9, we have a seismic line which passes through a wellbore. We see
depicted a fault moving through the area. This fault was anticipated to have 5000 feet of
throw. It was also anticipated in the vicinity of our critical casing window. As we analyze
the problem in Figure 4L.10, we see that once we set surface casing, we anticipate a
leak off or fracture gradient of about 14.2 pound per gallon mud weight equivalent. As
we drill down to the fault and cross it, if the first think we see is permeability, we would
see pore pressures which exceed our surface casing shoe integrity. This can lead to a
very hazardous situation and has a high risk for an underground blowout at the very
least.
The only way to safely drill a situation like this is to stay well above the anticipated fault
zone and set a protective string of casing. This is illustrated with an 11-7/8 inch
protective casing string being proposed at 9900 feet. This would provide us a much
higher leak off test at that point. If we then drill across the fault and encounter
permeability, we could certainly handle the pore pressures to be seen. However, not all
our problems would be eliminated. We must take into account again, Figure 4L.11,
differential sticking considerations.
In Figure 4L.12, we have plotted the differential pressures to be expected with respect to
depth as mud weight is raised in response to the pore pressures seen. We see that once
we set our 11-7/8” casing, our differential pressures drop down due to casing off all the
nine pound sands up the hole. We still have 11 pound per gallon formations exposed
immediately below the 11-7/8” casing seat.
As we drill across the fault, if the first thing we see is permeability and have to weight up,
our differential pressures are likely to exceed those we can safely handle. We still have
a high risk well from a differential sticking standpoint. If we do not see permeability
immediately, we still have a need for setting another intermediate string of 9-5/8” casing.
Once we do see permeability, we will have to raise mud weights beyond our safe
differential limitation. When we have faults at casing points, in a situation such as this,
we still have a high risk well. When we cross the fault, we may exceed differential limits
if the first thing we see is permeability. Setting the extra string of pipe merely the safety
hazard or the risk of an underground blowout.
Figures 4L.13, 4L.14, and 4L.15 illustrate how this casing program looks with one string
of casing above the fault, one immediately crossing it, and of course, one at authorized
depth.
In Figure 4L.16, we have a structured map. Assume that no wells have been drilled in
the fault block illustrated. We have at about right center a structure or platform from
which wells are drilled. It is desired to see this formation at a structural position which is
below the peak of the structural high to stay down in the oil leg. We propose a wellbore
in the direction of north, 66 degrees west.
In Figures 4L.17 and 4L.18, we have 1” log sections on a couple of wells in the area
which illustrate two things. In Figure 4L.17, we have a transition zone between 9100 feet
and 9200 feet where the pore pressure increase from 11.4 ppg to 13.2 ppg. As we look
deeper, we see a D5 sand which has a 14.5 pore pressure. In Figure 4L.18, we have
this same transition zone between 9300 feet and 9400 feet. We also note additional
sands above the D5 sand in this field such as a D1, D2, and D3 sands.
Figures 4L.19 through 4L.22 are the fault plane map of the “B” fault, and additional
structure maps of other formations in the area from which a cross section can be drawn.
If we draw a cross section through the proposed plane of North 66 degrees West, we get
Figure 4L.23, where we have faulted out our transition zone. We see the D5 sand shortly
below the cut, but are aware that there are several other possible sands above it, the D1
through D4 sands. We could cross this fault and not see our transition zone and
encounter permeability first thing.
If we go back to Figure 4L.16, and if satisfied with equal structural position, such as
illustrated by the directional heading of North 41 degrees West, we can cut this fault at a
shallower depth. As we draw a cross section in this plane, we arrive at Figure 4L.24.
We cross the fault at a much shallower depth and we should see the transition zone in
the drilling of our wellbore and can stop the well in this transition zone by moving the fault
away from the critical casing point. This is one of the values of drawing cross sections.
Having a cross section through the plane of the well we can see where we anticipate
faults, and what we can do about moving the wellbore or rotating the wellbore to move
that fault away from the casing point. This makes the well much more economical to drill
and much safer.
1. INTRODUCTION
In Figure 4M.2, we have a seismic section which illustrates such an effect. We can see
in Figure 4M.2, that as we move from the salt dome to the right that our formations are
expanding. Formations will be seen at deeper depths and much thicker. The importance
is in correlating and nailing down transition zones or casing points. Being aware of the
much longer sections that would be seen in proposed wells with respect to offsets will be
essential.
In Figure 4M.3, we have an ROP plot for a well which was drilled into a transition zone.
We see a hard spot on top of the transition at 10,100 feet followed by a drill off or
increase in penetration rate. A 1” electric log on this well appears in Figure 4M.4. Note
the reduction in resistivity and conductivity at the bottom of the hole.
In Figure 4M.5, we have replaced the conductivity curve with the ROP plot. Note the
close agreement between the two.
In Figure 4M.6, we have an ROP for an additional well in this block. As we move across
the block, we have a much longer section of hole where the resistivity and conductivity
drop off, or thicker sections in this part of the field, as in Figure 4M.7.
In Figure 4M.8, the ROP plot is spliced in the place of the conductivity curve. The reason
for the lack of drill off between 9,900 feet and 10,200 feet on the ROP plot is in
anticipating increases in pore pressure in this transition zone, increases in mud weight
are made. In an adjacent block, in Figure 4M.9, we have an ROP plot for a well drilled.
We see a much thicker hard spot in the vicinity of 11,900 feet followed by some
additional sands.
In Figure 4M.11 appears the ROP plot replacing the conductivity curve next to the
electric log. We see a much thicker hard spot in the vicinity of 9400 feet followed by
much thicker sands below it.
Note the much thicker shale section and the much thicker developed sands below the
hard drilling in Figure 4M.13. Figure 4M.14 illustrates the ROP plot replacing the
conductivity curve.
1. INTRODUCTION
After building an accurate geological model, consisting of an accurate pore pressure and
fracture gradient profile, it is now possible to select optimum, casing setting depths.
Considering the fact that the cost of casing and tubing is the greatest tangible
expenditure related to the overall drilling AFE, it logically follows that the depths to which
these tubulars are to be set has a major impact on the total well cost. This scenario
obviously points to the need for a sound engineering approach to casing point selection.
This section of the Modeling Manual will elaborate on the methodology associated with
this task. It should be kept in mind that even though all wells are not planned and
designed based on abnormal pore pressure, the building of, and utilization of a model
will, even in a normal pressured environment, provide clues to potential drilling problems,
anticipated wellbore instability areas, and ultimately casing setting depths.
2. GENERAL COMMENTS
The cost effective and optimum selection of casing setting depths is on a fundamental
level, a delicate balancing act whereby the well planner considers the relationship of a
proposed setting point to its pore pressure and fracture gradient environment plus, and
once again relative to a proposed setting point, the development of differential pressure
across any exposed sand, as well as considering mechanical risks. To say all of this in a
simpler way, we wish to designate a casing setting depth which does not violate
estimated fracture pressure limitations and also does not create a differential pressure
magnitude which increases the probability of differential sticking beyond a reasonable
level, as well as avoid high mechanical risk situations.
The later two criteria are obviously directly related to four sectors: the drilling fluid
density in the well at the time of running casing or drilling to the casing point, the true
vertical depth of the exposed sands, the estimated permeability of those sands, and
finally the degree of angular inclination projected for the well. To further complicate this
scenario, this model relative to its proposed setting depths is only a tool. The model is
always dynamic; certain wellbore and rig conditions will warrant changes in the proposed
setting point. The expertise required to know when to stop drilling and set pipe, or when
to continue drilling to push the point deeper, is not clearly defined by engineering or
planning criteria.
All technology that allows us to accurately determine casing setting point is relative to our
degree of accuracy in predicting pore pressures and fracture pressures. Time and effort
must be allocated to developing this skill.
Two major considerations will generally control the setting depths of surface casing. The
first of these is usually a federal, state or local regulation that will dictate a minimum
setting depth; these regulations are generally related to protecting the shallow sub-
surface environment. Exceptions to these regulations may be applied for and obtained.
The second consideration in picking surface casing setting depth, is that it should always
be set in a consolidated shale interval when practical. In other words, we should not set
surface easing until we have penetrated sediment where grain-to-grain contact has been
established. The grain-to-grain contact will provide a reasonable degree of formation
competency at the setting depth. This will provide a starting point for determining the
depth at which the intermediate casing can be set without creating a loss circulation
problem.
In addition, the observation and/or evaluation of leak-off test results is highly interpretive,
at best. For these reasons, relying solely on field leak-off data to establish a reliable
fracture pressure threshold is often a very costly practice. The purpose of this waiting is
not to expound on the short comings of field leak-off values. It suffices to say that at the
very best, the leak-off test values must be considered as tools only. Their application
relative to selecting casing points must be evaluated based on experience and
consistency of methodology.
Care must be taken in determining the magnitude of the fracture pressure limitation of
the surface casing shoe. The value will almost always play at least an indirect role in
determining the intermediate casing setting depth. Occasionally, its magnitude will be
the sole determining factor. Logic should, therefore, indicate that certain circumstances
will dictate a deepening of the surface casing setting depth in order to achieve a higher
formation integrity and, therefore, a greater fracture pressure threshold.
In general terms, four controlling parameters will interact to determine the safest and
most economical depth at which intermediate casing should be set. Even though these
controlling parameters are considered singularly, the final setting depth is the result of a
detailed analysis involving all four. Each will be discussed here, followed by a description
of this analysis.
As described in the previous section, the magnitude of the fracture pressure, or gradient
at the surface casing setting depth will, at least tentatively, provide a threshold drilling
fluid density above which the formation at the surface shoe will fracture. The fluid density
threshold correlates directly with a true vertical depth, as controlled by a plot of pore
pressure versus depth. This observation is usually the first step in selecting an
intermediate casing setting depth, and is considered to be only tentative.
The second parameter to consider is the probability of pipe differential sticking, should in
fact the previously determined tentative depth be utilized. Experience, and three major
studies conducted by the MMS and two major oil companies, provide us with differential
sticking guidelines, as previously discussed in this manual.
The tentative intermediate setting depth, that was previously determined relative to the
anticipate fracture pressure limitation at the surface casing setting depth, is now
evaluated using the differential pressure criteria. To verbalize the technique, the fluid
density in which the intermediate casing string well be run is applied across the deepest,
normally pressured sand. The differential pressure, given this environment, is calculated
by taking the difference between the drilling fluid density in the well and the pore fluid
density under normal pressure conditions, in the deepest normally pressured sand. This
difference is then multiplied by the pressure gradient constant (0.052) and by the true
vertical depth of the sand under consideration.
This product generates a differential pressure magnitude in psi. If the result of this
calculation is greater than 2000 psi for a straight hole, serious consideration must be
given to raising the tentative intermediate setting point to a shallower depth. Should the
above calculation generate a differential pressure value that is considerable less than
2000 psi, from a probability of sticking standpoint, lowering or deepening the tentative
intermediate setting point might be justified. It must be remembered, however, that the
initial setting depth was determined based on the magnitude of the anticipated fracture
pressure threshold at the surface casing shoe.
Assuming that this pressure threshold was accurately determined, deepening of the
intermediate string, in theory, is not possible. Generally, without violating engineering
principles, under these conditions the previously determined intermediate setting depth
would be utilized, or consideration might be given to deepening the surface casing setting
depth, thereby allowing a deeper intermediate setting depth to be chosen. A final
comment is necessary in relationship to differential sticking considerations. Obviously,
as the wellbore inclination departs from the vertical, differential sticking becomes more of
a concern.
After adjusting the intermediate setting depth for differential pressure considerations, the
third parameter is considered. The well is viewed from the proposed intermediate setting
point to the total depth point. It is then ascertained if the well can be drilled to TD without
setting a second intermediate string of casing. In many cases, the only point of
consideration for the lower most hole segment is whether or not the fracture pressure
limitation at the intermediate casing shoe is of sufficient magnitude so that it will support
the maximum drilling fluid density that the well will utilize. However, differential pressure
limitations need be analyzed in this hole segment as well. Should this criteria be
satisfied, the previously established intermediate setting depth is sound and the well
should be drilled below the intermediate shoe to TD without severe problems.
If, however, the fracture pressure threshold at the intermediate casing shoe is not
sufficient to support drilling fluid loads to TD, several choices are available. A second
long intermediate string may be planned to a depth, so that the bottom hold segment can
be drilled safely, or a drilling liner may be set inside of the primary intermediate string.
The drilling liner scenario will obviously require design modifications in the primary
intermediate string due to the fact that it will have to support loads generated due to
deeper drilling.
One additional option is available which will indirectly allow for deepening of the primary
intermediate string. Under certain circumstance, and assuming compliance with local,
state and federal regulations, it may be possible to deepen the setting depth for the
surface casing. Under normal geological development, increasing the depth of
investigation will increase the fracture pressure threshold. This increase will allow for a
deeper intermediate setting depth which may allow the bottom segment of the well to be
drilled without any major problems related to fracture pressures. It must be remembered
that differential pressure across the deepest normally pressured sand may still be the
controlling factor and either limit or negate entirely the possibility of deepening the
primary intermediate casing string.
Threshold differential pressures, as previously defined will set the depth limitations. The
tentatively chosen intermediate depth will then be adjusted if needed. After establishing
the adjusted setting depth, the bottom segment of the well below the intermediate setting
depth is examined for drillability relative to drilling fluid density requirements and the
fracture pressure threshold at the proposed intermediate casing shoe. Should there not
be enough formation integrity at this proposed shoe depth, several scenarios for
additional strings of casing or deepening possibilities must be considered. Finally,
adjustments need be evaluated based on mechanical risks.
Logic, therefore, implies that utilizing a kick tolerance will have the affect of raising the
previously determined intermediate setting depth. The specific true vertical depth, as
determined by utilizing a kick tolerance, is located at that point below which a 0.5 ppg
kick would in theory, fracture the previous casing shoe. There again, be reminded that
kick tolerance will be effected by influx volume, open hole length, casing points, etc. The
actual magnitude of the kick is based on the well planner’s prediction of the formation
pore pressure plus an 0.5 ppg increase.
Philosophically, this line of thinking proposes that the initial pore pressure prediction was
unsolved. Should the accuracy of the initial geological model (pore pressure prediction)
be somewhat suspect, and it is felt that a kick tolerance should be built into the well plan
as a contingency, it is recommended that an intermediate casing setting depth be
determined as per normal engineering practices (no kick tolerance), and a shallower
depth be defined as dictated by the kick tolerance.
Effectively, this well design provides a setting depth window, in which casing can be set
with reasonable assurance of safety and cost effectiveness. It must be remembered that
all well planning work serves only as a guide. Real time analysis of drilling progress and
efficiency will either verify the reliability of the planning work, or indicate a need to make
modifications.
It should be apparent that the reliability of pore pressure and fracture pressure prediction
is fundamental to optimum casing point selection. Accurate modeling is the key to safe
drilling practices, and increasing drilling efficiency.
5. EXAMPLE
The following section described the process of selecting an intermediate casing setting
point, utilizing a simple example. Figure 4N.2 is the starting point. It illustrates the
predicted formation pore pressure and the predicted formation fracture pressure. Upon
examination of this predictive plot, several points should be noted. Formation pore
pressure makes a transition from normal (9.0 ppg equivalent in this example) to abnormal
at approximately 8000 feet true vertical depth. The pore pressure increases quite rapidly
and continuously until reaching its maximum value of 14.5 ppg equivalent at a true
vertical depth of 12,100 feet.
The fracture gradient plot also illustrates an increasing trend. However, as expected, the
rate of increase in formation integrity is not parallel to that of pore pressure. The
operating “window” is becoming increasingly smaller with depth.
Note that the position of the deepest normally pressured (9.0 ppg equivalent) sand is
marked at 7700 feet true vertical depth.
We may now proceed with the design. The first step is to add a layer of safety margin on
the two predictive curves. This is done first by imposing an over-balanced drilling fluid
density schedule on the pore pressure estimation, and secondly, by underestimating the
predicted fracture gradient. Typically 0.3 ppg equivalent is used for the fluid density
over-balance, Figure 4N.3, and 0.5 ppg equivalent is used as a safety margin relative to
the predicted fracture pressure, Figure 4N.4 Notice that this has the effect of reducing
even further the operating window. Note that at TD the fluid density will be 15.0 ppg and
the fracture gradient, with design constraints is 17.0 ppg.
Now, it is possible to begin considering casing setting depths. Assume surface casing
will be set at 4000 feet, the fracture gradient at that depth will generally indicate the
maximum fluid density which can be utilized before that point (4000 feet) must be “cased
off”. The calculated fracture gradient at that point is 14.2 ppg. Remember, however, we
have given ourselves a safety margin of 0.5 ppg. This reduces the applicable fracture
pressure from 14.2 ppg to 13.7 ppg. How deep can we drill before the fluid density in the
well reaches 13.7 ppg? Figure 4N.5 illustrates a graphical approach for answering this
question.
If we work from bottom up as in Figure 4N.5, we see that a minimum of 9000 feet of
intermediate casing is re1uired to reach total depth, however, if we analyze further, we
see that we can drill to approximately 10,400 feet before a 13.7 ppg mud is needed.
Analyzing differentials across the last 9.0 ppg sand at 7700 feet, we see that we can live
with as high as a 14.0 ppg mud. We, therefore, have a casing point range for this well.
We tentatively have a minimum intermediate shoe requirement of 9000 feet and finally a
maximum allowable shoe depth of 10,400 feet.
The example proposes a well configuration which by our best engineering skills is sound.
The well is safe in terms of the relationship between estimated pore pressure and
formation integrity. The design also creates an environment where the probability of
differential sticking is quite low. If our information and its interpretation are correct, we
should be able to drill this well, as designed with minimum problems and maximum cost
effectively.
Finally, at its very best, this is only a plan. It guides the man drilling the well, but does
not dictate to him. The ability of modify this plan as the well guides the well dictates, is
essential to cost effective, optimize drilling programs. Throughout the entire process of
well planning or geological modeling, it is essential to remember that all of the
engineering work put into such a design is only as valid as the data used to generate the
initial estimation of pore pressure and fracture gradients. The quantity, quality and
utilization of valid geological data is the foundation for accurate well planning.
6. ADDITIONAL EXAMPLES
After the pore pressure and fracture gradient prediction process is complete, barring any
other modeling considerations, we determine where each casing point will be and
consequently how many strings, and their sizes, will be required. The example in Figure
4N.6 is complicated by the addition of a fault, the crossing of which results in a rapid pore
pressure increase. This necessitates a safety string well above the fault and another
upon crossing.
In the scenario of Figure 4N.7, three intermediate strings of casing are required.
In Figure 4N.8, two intermediate strings are required to drill through a small interval.
In the case of Figure 4N.9, its questionable as to the need of the third intermediate so
close to total depth. It would be a logical risk to plan the well without it.
Figure 4N.10 illustrates a case where intermediate casing is required due to differential
pressure limitations.
When abnormal pressures are seen to develop at very shallow depth and continue to
build, many strings of casing can be required to drill in these environments as in Figure
4N.12.
When pore pressures approach fracture gradients, we approach a point beyond which
we may not be able to drill as in Figure 4N.13.
The directional considerations of a well are greatly affected by the modeling process. In
the illustration of Figure 4N.14, if we began to build angle at the prescribed kick off point
of 4450 feet, we would not fully develop our angle until below the next required casing
point.
It, therefore, becomes necessary to determine the pressures to be seen prior to planning
the directional requirements as in Figure 4N.15. Not only are the directional concerns
dependent on the modeling process, but so too are all other drilling considerations.
The second intermediate string in the example of Figure 4N.16, would be required rather
than attempting to drill a long interval so close to balanced.
In Figure 4N.17, pore pressures do not develop to extremely high magnitudes in this
example. The selection of the intermediate point should be to minimize differentials in
each hold section.
Figure 4N.13
Figure 4N.14
Figure 4N.15
Figure 4N.16
Figure 4N.17
From what we may have seen so far, it is apparent the geological needs of drilling are
great in well planning. It is necessary for a detailed informational analysis, not only in the
planning phase, but during the executable phase. Logs need be analyzed as run to
confirm predicted trends as in Figure 4O.1.
The location of transition zones, as in Figure 4O.2, need to be verified by log analysis
prior to running casing.
Pore pressures need to be confirmed from log analysis comparisons of normal trends vs.
abnormal readings to verify or determine if our anticipations are correct as in Figure
4O.3.
Since our initial geological concepts, such as the structure map illustrated in Figure 4O.4,
were an interpretation of the available data prior to drilling, as the geologist modifies his
perceptions of the structure from information gathered while drilling, the drilling man need
be aware of modifications which may alter his anticipated formation tops.
Anticipated faults, as in Figure 4O.5, may not be encountered or may have a significantly
different throw than originally estimated.
For example, in the case of the well in Figure 4O.7, formations were anticipated to be
seen approximately 1000 feet updip of the offset well from the best geological information
available in the planning phase. In actuality, formations were found to be only 440 feet
updip from the offset in the drilling process resulting in pore pressures being lower than
anticipated.
Figure 4O.8 illustrates another case of modifying the actual casing program during the
course of drilling from that of the plan. Note the elimination of the second intermediate
string of casing.
The need of this second string was based on differential pressure limitations below the
first intermediate string as illustrated in Figure 4O.9. Permeable sands were anticipated
below this string based on the offset wells.
However, during the course of drilling the well, a correlative ROP (rate of penetration)
plot, Figure 4O.10 indicates these anticipated sands to be shaled-out between the depth
of 12,500 and 12,850 with respect to the offset well.
If permeable sands are not exposed, differential sticking cannot occur. Consequently,
mud weight limitations can be increased as in Figure 4O.11, which in the case of this
well, allowed for the elimination of an intermediate string of casing.
In Figure 4O.12, we can see that actual pore pressures encountered were very similar to
those anticipated. So indeed, the lack of the permeable sands was the factor enabling
the elimination of the string.
Figure 4O.13 is an example of a well where the casing program remained the same as
proposed, however, the intermediate setting depth was somewhat shallower than
planned.
During the course of drilling this well, upon reaching a depth of approximately 7500 feet,
it became apparent that formations were coming in high as in Figure 40.14.
Realizing this, the pore pressures predicted were adjusted upward approximately 650
feet as in Figure 4O.15. Upon modifying the analysis at midpoint of the well, it was
apparent that the casing program was adequate, but merely required a shallower setting
depth. Upon adjustment, as can be seen here, the prediction became an accurate one.
One must think of the modeling and predictive process as a dynamic one. One which
changes as new information is gained during the course of drilling.
The dynamics of the drilling model, or need for modifications, are due to the interpretive
nature of the information. The greater the amount and quality of information, the more
accurate the model, but new information gained during the drilling process always adds a
new piece to the puzzle.
One interpretation of the structure is indicated as one structural high against the fault as
in Figure 4P.2.
Another, equally valid, interpretation of the same data is with two structural highs against
the same fault as in Figure 4P.3.
If we look at a plane through the high of the first interpretation, Figure 4P.4.
However, if we look at a plane through a high of the second interpretation, Figure 4P.6.
We would perceive beds which dip significantly less than before as in Figure 4P.7.
1. THE BASICS
In Figure 4Q.1, we have a Biostratigraphic Nomenclature Chart for Chevron. This can be
utilized to determine geological ages from Chevron paleo data.
Figure 4Q.2 relates geological time to historical events. Please refer to the following
pages for the Figures.
Fault terminology appears in Figure 4Q.3. Nearly all anticlinal reservoirs are broken by
faults. Usually they are vertical and strike at an angle of about 70 degrees to the axis of
the anticline.
The dynamics of river transport are illustrated in Figure 4Q.4. These are known as fluvial
environments. When a river is no longer digging the bottom of its bed, it tends to flow in
great sweeping curves called meanders.
As the meanders migrate downstream, they form multiple point bars, as illustrated in
Figure 4Q.6.
A typical channel sand deposit is illustrated in Figure 4Q.7. If the river abandons the
channel, as by the cutoff of a meander or upstream diversion, the current finally stops
completely. Eventually, the abandoned sand-filled channel is buried by flood-plain
deposits.
Channel deposits can be recognized by their erosional base, which truncates older
stratified deposits and causes an abrupt change in lithology to a coarse sand. In the
lower part are frequent chunks of clay, apparently pieces of the stream bank which fell in.
If these chunks are large enough to exceed the diameter of the wellbore, these sands
may appear to have shale streaks separating layers of sand or as vertical permeability
barriers on an electric log. Vertical permeability may exist, however, beyond the
wellbore.
Channel sands can often be recognized on the electric log using the gamma-ray SP, and
short-spacing resistivity curves. The base is abrupt. Usually, both SP and resistivity are
a maximum in the lower layer because the sand contains less clay and has a higher
porosity. The overlying beds contain more clay, both interstitial and in laminae, so both
SP and resistivity decrease in amplitude and become serrated near the top.
The electric log response of a channel sandstone appears in Figure 4Q.8. Water velocity
in stream channels varies, and when it is reduced, causes setting of the heaviest and
largest components first. Continued reduction of velocity causes progressively fine
materials to be deposited over the coarse beds. Typical electric-log response to channel
sands indicates an increase in SP in the bottom of a channel sand, which can
occasionally imply increased permeability. Knowing the environment to be that of
channel sands should imply anticipation of differential sticking to be more likely at the
base of the sands than at the top due to the increasing permeability as we drill through
the sand. If the hole is acting sticky at the top of the sand, differential sticking is likely at
the base and vice versa.
A typical beach or barrier bar sand appears in Figure 4Q.9. The waves of the ocean
continually strike the shore. At the front of a delta, the waves winnow the clay from the
sand, building up a beach of clean sand. Such beaches are often buried by layers of
mud as the delta builds outward. They thus become completely enclosed in shale and
form stratigraphic traps for oil. The rush of the waves and along-shore currents smooth
out the coastline so that beaches and barrier islands are often nearly straight. Behind
the barrier island is a quiet lagoon where both sand and mud are deposited. The grain
size and cleanness of the sand thus decrease away from the ocean toward the land.
Thus, as we head inward from the beach area, significant reductions in permeability can
be expected.
An electric log response of a beach deposit appears in Figure 4Q.10. Electric log
response of beach deposits is the inverse of that for stream channel sands. Beach
sands are deposited upon fine-grained sediments that have little porosity and reduced
SP and resistivity response. Therefore, permeabilities are likely to be higher at the top
with a gradual reduction as we reach the base. If differential sticking is not a problem at
the top of the sand, it is not likely to occur as we drill deeper to the base.
A typical beach-type sand appears in Figure 4Q.12. Beaches and barrier-island sand
bodies have the cleanest and coarsest sand in the upper part, where it is most washed
by the waves. The sand is often stirred up by burrowing organisms such as clams. The
seaward pinch-out is abrupt, smooth, and straight. The landward side toward the lagoon
tends to be transitional, the sands becoming dirty and inter-fingering with the shale
(muds) of the lagoon. The lagoon side is irregular. As we head seaward, toward the
pinch-out, the sands thin and permeabilities decline. If differential sticking is not a
problem in the vicinity of the beach, it becomes less likely as we head farther toward the
sea.
Some typical representations of sphericity and roundness in sand grains appear in Figure
4Q.14.
Some of the basic terminology and structure of clay minerals appears in Figure 4Q.15.
There are four principal types of clay minerals. The kaolinites consist of one silica
tetrahedral layer and one alumina octahedral layer linked together. They adsorb water
only around the edges, not between the layers; they do not swell. The chemical
composition of kaolinite is (OH)8Al4Si4O10.
The montmorillonite group, now called smectite, consists basically of three layers: one
octahedral alumina layer with a tetrahedral silica layer on each side of it. Water can get
in between the tetrahedral layers. With one layer of water, the c-dimension of the lattice
is 9.6 Angstrom units, and with more water, the c-dimension increases to 21.4 Angstrom
units. Smectite, therefore, swells when it is placed in water. It also tends to disperse into
tiny platelets. The chemical composition of smectite is (OH)4Al4Si8O10nH20. Potassium
ions fit between the hexagonal ring of the silica tetrahedra. When they do so, they bind
the layers together so they cannot swell. These clay minerals are called illite. Some iron
and magnesium are also present in illite.
The chlorites are similar to illite but contain iron and magnesium. Other types of clay
minerals also occur, but nowhere in abundance.
Illite is related to smectite. When deposited in the pores, it often has a very open
honeycomb or fibrous structure that causes large loss of effective porosity. It does not
swell as much with fresh water, but the particles do migrate to pore throats, resulting in a
loss of permeability.
There are certain effects of drilling mud on sandstone to be considered. When drilling
over-balanced, a filter cake develops on the face of a permeable sandstone. In order to
minimize invasion of mud and mud filtrate, steps are taken to decrease the permeability
of the filter cake. Most of the solids in the mud remain in the filter cake and do not enter
the sand pores. However, a small amount of fine particles does enter and tends to plug
the pores in the vicinity of the wellbore.
The filtrate from the drilling mud enters the sand and often invades it for a distance of
several feet. If the sand contains swelling clays, they may originally have been
flocculated in the salty, connate pore water. When this is displaced by fresh water,
swelling occurs. Certain chemicals put in the drilling mud to reduce its water loss will
deflocculate the clays in the pores. Even if there is no swelling, the flow of water may
pick up and move the delicate authigenic clay particles, causing blocking.
If there is oil in the rock, the water may form stable emulsions or asphalt films that block
the pores. If the chemicals in the drilling mud react with the compounds in the pore water
to form precipitates, these will also cause blocking. The surfactants in the mud may
cause the oil to be displaced so completely that the permeability of the sand to oil is
greatly decreased.
Figure 4Q.16
A comparative illustration of permeability grain size and shape appear in Figure 4Q.17.
Grains oriented in one direction can increase rock permeability parallel to their long axes
and reduce it normal to their long axes. This is particularly true where small, flat shale
grains provide no permeability normal to the fissility of the shale, but can allow the lateral
movement of fluids along it during compaction.
Figure 4Q.18
An abundant sedimentary rock that often contains oil is limestone. Sometimes the
limestone contains substantial amount of magnesium, replacing calcium, and it then
becomes dolomite. It has become customary in the oil business to call both limestone
and dolomite carbonates to avoid making a distinction.
It has been estimated that about half the world’s oil reserves are in carbonates, although
there are numerically fewer carbonate than sandstone reservoirs outside the Middle East.
Carbonates differ in many respects from sandstones. They are mostly formed from the
remains of animals (shellfish) and plants (algae); they are, therefore, found in nearly the
same place where they originated and were not transported and then deposited like
sandstones.
The matrix is lime of clay-particle size (lime mud). It is called micrite. The clear
secondary calcite cement is called sparite.
Thus, a rock consisting mainly of clear secondary calcite with intraclast grains would be
called “intrasparite”. A rock consisting mainly of micrite (lime mud) with grains consisting
of broken shell fragments would be called “biomicrite”. Biomicrite and pelmicrite are the
most common limestone types. These eight types are shown diagrammatically in Figure
4Q.19.
Besides these eight combinations, there are some limestones consisting only of micrite
and some consisting of the remains of upstanding reef-building organisms. So there are
ten types of limestones in all.
The types of grains typically seen in limestones appear in Figure 4Q.20. The type of
grains is considered significant. Five types of grains: detrital “intraclasts”; skeletal “bio”;
pellets “pellets”; lumps (irregular clumps; and coated grains “ooliths”.
The carbonate depositional environment and characteristic rock types appear in Figure
4Q.21. Many, if not most, ancient carbonates were deposited simultaneously in three
different macro-environments - shelf, slope, and basin.
Shelf
The shelf environment consists of broad, shallow seas, mostly less than 100 feet of
water. Currents are weak, so generally lime mud has been deposited. Usually, there is
enough current or wave motion to keep the water oxygenated. Scatter isolated coral
head or larger patch reefs are common. Sometimes in mud banks, oxygen is used up
and organic matter is preserved. However, if the water does stay oxygenated, aerobic
bacteria act upon the organic matter and destroy it. In such instances, connection and
background gasses would not be reliable indicators of pressure.
Slope deposits
The material on the slope consists of lime sands and blocks that have been broken off
the reef by waves and deposited in strata with an initial dip. They are called reef talus
and sometimes form excellent reservoirs.
Basin deposits
The material in the basin is fine grained, usually lime mud. Normally, it does not have
sufficient permeability to produce hydrocarbons. In a few places, chalk has accumulated,
formed from the tiny shells of algae called coccoliths. They have considerable porosity
but very low permeability. The basinal carbonates often grade laterally into shale. In the
case of epi-continental basins, it often happens that there is little circulation of the water
in the deeper parts of the basins. Organic matter is preserved because not enough
oxygen is brought in to destroy it. Occasionally such deposits become highly organic and
may become source rocks of hydrocarbons. Connection and background gasses can be
reliable indicators of increasing pressure in such environments, however, low
permeability is generally likely. Such formations typically provide low volumes and rates
of flow when drilled under-balanced.
Limestones are often partially or completely changed to dolomite. Dolomite has the
composition CaMgCO3 and it is crystallographically similar to calcite. However, it has
greater density, less solubility in water, less ductility, and more brittleness.
Dolomites are always found in the shelf environment, near in the deep basin. They are
commonly associated with evaporites (salt and especially gypsum and anhydrite) and
with stromatolites, which are layers of algael mats.
There is no doubt that dolomites generally are more porous and permeable than
limestones.
An illustration of fracture porosity and tension type fractures appears in Figure 4Q.22.
When sandstones are sharply folded, the beds slide over each other; but when
carbonates are folded or faulted, they develop fractures. If the stress environment is
compressional, as it often is along faults and sharp folds, the fractures may be tightly
closed. In this case the porosity due to the fractures will be negligible, and even the
permeability will not be very great. Much more prolific are fractures resulting from
tension. Over an anticlinal told, tensional fractures are perpendicular to the bedding and
either parallel to the axis of the fold or, more commonly, at right angles to it. If fractures
are anticipated in a tension environment, lost circulation is possible.
Fractures usually have a volume of less than 1 percent of the rock, while the matrix may
have a porosity of 5 to 10 percent. Consequently, fractures contribute much more to the
permeability of a reservoir than to its porosity.
An illustration of chalk appears in Figure 4Q.23. Chalk is formed from microscopic algae
called coccoliths. Their tiny ring-shaped shells, 1-20 um in diameter, were made of
calcite. These settled to the bottom, forming a thickness of chalk. Chalk is an
exceedingly pure carbonate rock, soft, white and porous, that contains very little
terrigenous material. Locally, the chalk may contain small amounts of shell debris.
Nodules of chert are common. When first deposited on the sea bottom, the porosity of
the shells is 70%. With increasing depth of burial, porosity is lost rapidly. Some of the
loss is due to mechanical compaction and some to chemical cementation. At a depth of
burial of 1,000 m, the porosity of most chalk is reduced to about 35%, at 000 m to 15%,
and at 3 km practically to zero. The permeability of chalk is typically very low because
the particle size is so fine. It decreases from about 10 md at 40% porosity to 0.1 md at
10%. Chalk is also a good seismic reflector.
Quite simply put, the origin of reservoir pressures appears in Figure 4Q.24. In the cross-
section, the aquifer outcrops at A. There are three oil fields in the same aquifer: B, C,
and D. In each field, the pressure will be that necessary to sustain a column of water to
the elevation of the outcrop.
The low-pressure reservoirs are in well-consolidated sediments which have been uplifted
in the recent geologic past, and which are now undergoing erosion.
Figure 4Q.24b
AtPage
temperatures
Q - 26 of about 100°C or depths between 8,000 and 10,000 feet (2,500 and Rev. 4/24/90
3,000), smectite converts to illite. This involves a loss of interlattice water. If there is an
increase in specific volume of the water as it comes out of the smectite or if the smectite
CHEVRON DRILLING REFERENCE SERIES
VOLUME FOUR
PORE PRESSURE PREDICTION
Figure 4Q.24a
As deposition of the organic material takes place in the marine environment, it is buried
and protected by the clay and silt that accompanies it. This prevents decomposition of
the organic material and allows it to accumulate.
Shale and some carbonates contain organic material that bears hydrocarbons. Shale
and carbonate rocks of this type are not reservoir-type rocks and could be considered
ultimately to be source beds.
Diagenesis of source-type rocks eliminates some of these organic materials but allows
retention of residual amounts that are found in some quantities in most non-reservoir
rocks.
Therefore, if aerobic bacteria can act, there will be no shale or carbonate gas. If they
cannot act, there will be shale or carbonate gas. (Background or connection gas)
geologic past, and which are now undergoing erosion.
-6
A sandstone reservoir contracts elastically about 7 x 10 pore volumes per pore volume
-5
per psi (5 x 10 pv per kPa) as the internal pressure of the fluids is removed. This is an
elastic compression and should not be confused with compaction, which is irreversible.
The removal of overburden will cause an elastic dilation of the sandstone at about the
same rate. Shales appear to contract and dilate even more than sandstones. The
-6 -7
modulus of compressibility of water is about 3 x 10 volumes per volume per psi (4 x 10
vol/vol/kPa). Therefore, as overburden is removed, the pore volume dilates but the
interstitial water expands only about half as much as necessary to fill the new pore
volume thus created. Consequently, its pressure will drop.
For the reservoir to remain at subnormal pressure, we must assume that the overlying
shales are completely impermeable. Otherwise, the ground water would filter down and
pressure up the aquifer.
At temperatures of about 100°C or depths between 8,000 and 10,000 ft. (2,500 and
3,000), smectite converts to illite. This involves a loss of interlattice water. If there is an
increase in specific volume of the water as it comes out of the smectite, or if the smectite
loses volume, it could cause an increase in pore pressure.
If the pore pressure is higher than the mud pressure the shales tend to spall and cave
into the hole. It the pore water in the shales is saturated with methane, as is often the
case, it comes out of solution on the way up the hole and makes a strong indication in
the gas-measuring device in the mud logger.
Current information dealing with the organic origin of petroleum indicates that
hydrocarbons are products of altered organic material derived from microscopic plant and
animal life.
As deposition of the organic material takes place in the marine environment, it is buried
and protected by the clay and silt that accompanies it. This prevents decomposition of
the organic material and allows it to accumulate.
Conversion of the organic material takes place in the marine environment, it is buried and
protected by the clay and silt that accompanies it. This prevents decomposition of the
organic material and allows it to accumulate.
Shale and some carbonates contain organic material that bears hydrocarbons. Shale
and carbonate rocks of this type are not reservoir-type rocks and could be considered
ultimately to be source beds.
Diagenesis of source-type rocks eliminates some of these organic materials but allows
retention of residual amounts that are found in some quantities in most non-reservoir
rocks.
Therefore, if aerobic bacteria can act, there will be no shale or carbonate gas. If they
cannot act, there will be shale or carbonate gas (background or connection gas).
The vertical stresses existing in sediments appear in Figure 4Q.25. When the pressure
in the pore water approaches the weight of the over-burden, the overlying strata are
practically floating.
The weight of the overburden (S) is sustained by the stress in the skeleton of the solid
grains % and the pore pressure (p) in the interstitial fluids.
S = % + p
As p increases, % decreases and may become very small. That is, the solid skeleton is
supporting very little weight, and the overlying strata are floating. They can slide under
weak lateral forces, such as gravity sliding if the area is tectonically tilted.
Most, it not all, low-angle thrust faults probably take place in a zone of abnormally high
pressure. A seismic illustration of growth type faulting with featureless shale zones
appears in Figure 4Q.26. Seismic cross-sections clearly show the faults and dips of the
stratified beds. They also show where the shale has become chaotic below the fault
planes. Some of these featureless shale zones may be caused by diapirism deep below
the surface, while others may represent the toe zone of the slump block where the fault
emerges at the surface, part way down the continental slope,
Shale in the chaotic zone is under-compacted and contains fluids at pressures almost
equal to the weight of the overburden. An under-compacted bed is less dense than a
normally compacted bid. It, therefore, is unstable and has a tendency to be forced
upward.
1. BACKGROUND
We are currently studying the problem of borehole stability from several aspects at the
DTC; whether it is due to abnormally pressured shale, borehole inclination, clay
chemistry, etc., as well as pursuing any industry data or research on the subject. The
problem appears to be a very complex one with many contributing factors.
Chevron, through the DTC, has participated in studies evaluating the effects of various
mud additives on bore hole stability which have yielded many significant results. In
addition, Chevron participates in associated joint industry DEA projects. We have
developed numerous techniques within Chevron for the determination of formation pore
pressures and, in comparison to all others, find them to be the most accurate available in
the industry today.
We have investigated several Chevron wells where “borehole ballooning" has taken
place. When all the necessary data has been available for a complete analysis, we find
our problem primarily relates to a combination of a lack of detailed formation pore
pressure analysis, inadequate rock composition analysis, and excessive surge pressures
exerted on the borehole while tripping pipe.
The ballooning effects are generally seen in the deeper sections of our wellbores below
an intermediate string of casing. Once intermediate casing is set, we have a long
piston/cylinder relationship in place. Intermediate pipe is set in some mud weight and a
formation integrity test below the casing shoe is made. Generally the formation integrity
at this intermediate shoe is much higher, in mud weight equivalent, than the mud weight
casing was set in.
As drilling progresses to deeper depths and higher pressures are encountered, the mud
weight is increased to balance this higher pressure. As our mud weights increase, they
approach the magnitude of the intermediate shoe integrity. During a trip to change bits it
becomes very easy to initiate a fracture in the formations just below the casing shoe due
to the long piston/cylinder relationship we have between the drill string and the casing.
The combination of mud weight in the hole and the surge pressure created while tripping
in the hole can exceed the formation integrity, thus initiating a fracture in the formation.
Once a fracture has been initiated, extension of the fracture can be made at a lower
pressure, or at a lower mud weight equivalent. If the mud weight in the hole at the time
of fracture initiation is below the extension pressure, then the hole stays full at the time of
fracture initiation. As drilling progresses to even deeper depths and higher pressures are
encountered, the mud weight is raised further. We then get to a point where a
combination of the mud weight in the hole and the friction pressures due to circulation
exceed the fracture extension pressure and consequently we begin to lose mud while
drilling as mud begins to extend the fractures. When we stop circulation, the mud weight
alone, without the help of friction pressures, no longer exceeds fracture extension
pressure. The forces of the overburden now act in trying to force the fractures closed,
resulting in squeezing the mud back into the wellbore, or appearing to flow.
In many areas of the country we also encounter, as our drilling proceeds below a casing
shoe, naturally fractured formations. With these naturally fractured formations we can
induce the same phenomena as above.
When utilizing oil base drilling fluids, a close relationship exists between the required
water phase salinity of the mud, and the salinity of the formations drilled and formation
pore pressure. Exceeding the required water-phase salinity of the mud results in
chemically altering the formation rocks by drying due to osmotic forces. This chemical
alteration can result in subsequently weakening the rock and lead to a reduction in
fracture initiation pressure. A similar "ballooning" relationship can then be experienced
as previously described. This problem with oil based drilling fluids is further complicated
by the compressibility of the fluid yielding higher mud weight equivalents down hole than
seen at the surface.
It has been proposed to Chevron, by Mr. Gill, that borehole ballooning of the expanding
and contracting borehole theory, is the cause of our need of high mud weights in the
Mobile wells. It has been suggested that our Mobile wells are in actuality, normally
pressured. However, our in house techniques of analysis suggest the contrary. In the
drilling of the Mobile 861 No. 1, the Norphlet was encountered with a 16.7 ppg mud which
resulted in a kick and an underground blowout. Chevron's analysis of the pressures seen
indicated the Norphlet to be a 17.2 ppg pore pressure. In the Mobile 862 No 2, we
therefore, drilled the Norphlet with a 17.4 ppg mud. No problems were encountered and
the well was successfully drilled. By Mr. Gill's theory, we should have encountered
problems with a ballooning borehole but did not. Whenever we drill wells in an
abnormally pressured environment, as we weight up the drilling fluid to handle the higher
pressures seen, we expose the lower pressured formations up the hole to high
differential pressures. By Mr. Gill's theory this would always create an impossible
situation of flow from these shallower formations, making it impossible for us to drill in
abnormally pressured environments. However, this problem does not occur.
We at Chevron have visited with Mr. Gill on several occasions to discuss his theories.
He claims faster penetration rates to be the result of ballooning. However, on numerous
abnormally pressured wells we observe increases in penetration rates associated with
increasing pore pressure with no associated ballooning. He states that connection
gasses are never seen above 120 units. Many times we experience much higher values
than this in our operations in abnormally pressured environments with valid bottom hole
pressures recorded to verify the presence of abnormal pressure. Mr. Gill claims that
there are no abnormally pressured reservoirs, but he has no explanation for the
numerous production histories we have in such abnormally pressured environments. He
insists the Mobile wells we've drilled are normally pressured but is unable to explain the
abnormally pressured underground flow we experienced for months in the 861 No. 1 well.
Mr. Gill's paper was reviewed twice by two groups of Co-Technical editorials plus their
advisors, the SPE Editorial Review Committee. On both occasions his paper was
declined for publication because of lack of technical merit and composition.
Mr. Gill has approached some of our drilling superintendents in the past with his concepts
of the "ballooning" borehole which resulted in drilling abnormally pressured wells under
balanced, kicks, stuck pipe and lost hole. We can attribute several lost holes in our
operations due to employment of his theories. We see him and his theories as very
dangerous to the industry.
The solution to many of our instability problems can be solved through an increased
understanding of the environment we are about to drill. This can be developed through a
detailed pore pressure analysis, rock composition analysis via logs, mud logs, cores and
paleo data, and a better understanding of the structural relationship between wells
surface to total depth. This would require a change in philosophy, as a company, from
one of reactionary, and rushing our drilling prospects from every planning aspect, to that
of a detailed planning approach and project team analysis.
2. DISCUSSION
Numerous wells, where the so called "borehole ballooning" has taken place, were
evaluated in an effort to determine the mechanism by which we experience either one of
two phenomena. One being the gradual loss of drilling mud to the wellbore while drilling,
followed by a flow back of the mud whenever the pumps were stopped. The second
being the low volume, low rate influxes of formation water into the wellbore both while
drilling or with the pumps off, without any associated mud losses. Generally these low
rate influxes have been on the order of 1/4 to 1 barrel of influx per hour. When all the
necessary data has been available for a complete analysis, we find our problem primarily
relates to a combination of a lack of detailed formation pore pressure analysis,
inadequate rock composition analysis, and excessive surge pressures exerted on the
borehole while tripping pipe.
The ballooning effects which are characterized by the gradual mud loss while drilling, with
a flow back with the pump off, are generally seen in the deeper sections of our wellbores
below an intermediate string of casing. Once intermediate casing is set, we have a long
piston/cylinder relationship in place, with the casing acting as a cylinder and the drill
string acting as a piston. Intermediate pipe is set in some mud weight and a formation
integrity test below the casing shoe is made. Generally the formation integrity at this
intermediate shoe is much higher, in mud weight equivalent, than the mud weight casing
was set in.
As drilling progresses to deeper depths and higher pressures are encountered, the mud
weight is increased to balance this higher pressure. As our mud weights increase, they
approach the magnitude of the intermediate shoe integrity. During a trip to change bits it
becomes very easy to initiate a fracture in the formations just below the casing shoe due
to the long piston/cylinder relationship we have between the drill string and the casing, as
illustrated in Figure 4R.1.
In Figure 4Q.1 we have set casing in a transition zone at a point where formation pore
pressure is 12 ppg mud weight equivalent. Upon drilling out of this casing shoe, a
formation integrity test is taken yielding a 16.5 ppg equivalent. An illustration of the
recorded data during the integrity test appears in the illustration as well. Note the
deviation from the straight line trend at a value of 16.5 ppg equivalent mud weight
indicating leak off. Normally the test would have been concluded at the dashed line
before formation break down had occurred. Beyond the dashed line we illustrate
graphically as we continue with such a test and proceed to break down or fracture the
formation. Note that we arrive at a fracture extension pressure of 15.5 ppg equivalent
mud weight, somewhat below the leak off point. At this pressure, if constantly imposed
on the wellbore, we could continue to extend formation fractures well beyond the
wellbore as we pump fluid into the fracture.
The combination of mud weight in the hole and the surge pressure created while tripping
in the hole can exceed the formation integrity, thus initiating a fracture in the formation.
In this illustration the mud in the hole weighs 14.0 ppg due to currently drilling a formation
of 13.5 ppg pore pressure. A pore pressure profile vs. depth for this example appears in
the illustration as well. The surge pressure exerted at the shoe due to flipping the drill
string is 3.0 ppg equivalent. The combination of mud weight and surge pressure yields a
17.0 ppg equivalent mud weight at the casing shoe, thereby exceeding formation integrity
and initiating fractures. If the mud weight in the hole at the time of fracture initiation is
below the extension pressure, as it is in Figure 4R.1, then the hole stays full at the time
of fracture initiation.
Once a fracture has been initiated, extension of the fracture can be made at a lower
pressure, or at a lower mud weight equivalent. As drilling progresses to even deeper
depths and higher pressures are encountered, the mud weight is raised further, as in
Figure 4R.2. Here we have arrived at a formation of approximately 15.0 ppg pore
pressure, necessitating a mud weighing 15.3 ppg. The friction pressures in the annulus
due to mud circulation are equivalent to .4 ppg mud weight. The combination of mud
weight and friction pressures equates to a 15.7 ppg equivalent thereby exceeding the
fracture extension pressure of 15.5 ppg.
Figure 4R.1
3. On the same offsets as above, gamma ray and sonics displayed in the
following scales, one inch = 1000 feet and one inch = 100 feet (also
with gamma ray and sonic values displayed on a linear scale).
Figure 4R.2
4. A bulk density log from at least one of the offset wells above. (In
addition, it would be desirable for the log to be integrated to determine
the relationship of Overburden Gradient vs. Depth and have this
information graphically displayed.)
6. Scout tickets, mud logs, any drilling summaries for the offset wells
above.
7. Any virgin BHP and current BHP information from any formations for
the wells above.
Figure 4R.3
9. Half scale (one inch = .4 sec), interpreted seismic lines which tie the
proposed location to the offset wells above.
10. Seismic base map indicating location of offset wells, proposed location
and lines provided.
Figure 4R.4
12. An ITT curve on at least one of the offset wells which see abnormal
pressure (pore pressures higher than the normal fluid gradient), but
preferably two offsets. The ITT on any offsets should be displayed
with the sonic logs of these offsets for comparison. The vertical scale
should be 1 inch = 1000 feet, the horizontal scale should be two cycle,
semi-log, in micro-seconds per foot.
Figure 4R.5
We then get to a point where a combination of the mud weight in the hole and the friction
pressures due to circulation exceed the fracture extension pressure. Consequently, at a
mud weight below the formation integrity test, we 'mysteriously' begin to lose mud while
drilling as mud begins to extend the fractures.
When we stop circulation, the mud weight alone, without the help of friction pressures, no
longer exceeds fracture extension pressure as in Figure 4R.3. The forces of the
overburden now act in trying to force the fractures closed, resulting in squeezing the mud
back into the wellbore, or appearing to flow. We now have developed a ballooning
appearance with loss of mud to the wellbore while drilling, and flow back with the pumps
off.
In many areas of the country we also encounter, as our drilling proceeds below a casing
shoe, naturally fractured formations. With these naturally fractured formations we can
induce the same phenomena as above.
A similar occurrence exists when using oil based drilling fluids, however, an additional
fracture initiation mechanism also exists. When utilizing oil base drilling fluids, a close
relationship exists between the required water phase salinity of the mud, and the salinity
of the formations drilled and formation pore pressure. The water phase salinities
required for stability are illustrated in Figure 4R.4.
Exceeding the required water phase salinity of the mud results in chemically altering the
formation rocks by drying due to osmotic forces. This chemical alteration can result in
weakening the rock and can lead to a reduction in fracture initiation pressure. A similar
“ballooning" relationship can then be experienced as previously described. This problem
with oil based drilling fluids is further complicated by the compressibility of the fluid
yielding higher mud weight equivalents downhole than seen at the surface.
In all wells where flow was experienced without any mud losses to fractures, pore
pressure analyses have determined an under balanced condition to exist. In many low
permeability hard formations, we are able to drill under balanced with low rates and
volumes of flow into the wellbore. This is illustrated in the following two examples.
In Figure 4R.6 we have a graphical display of pore pressure and mud weight used vs.
depth for the Destin Dome 422 No. 1 well. Note from approximately 15,000 to 22,000
feet the mud weight used falls below pore pressure determinations. Throughout this
interval, while drilling the well, influxes of 1/2 to 3/4 barrels per hour of formation water
were encountered. As expected, this occurred due to an under balanced condition while
penetrating hard, low permeability formations.
Figure 4R.7 is a graphical display of pore pressure and mud weight used vs. depth for
the Mobile 862 No. 1 well. Note at a depth of 19,741, while tripping pipe out of the hole,
an influx of approximately 3/4 barrels per hour was encountered. This was due to
encountering a low permeability salt water formation under balanced at this point. After
the trip the mud weight was gradually raised to exceed the pore pressures and with a
15.8 ppg mud in the hole, no influx of fluid was experienced on subsequent trips.
In addition, many of our troublesome shales, we're finding, are abnormally pressured
although the surrounding permeable formations are normally pressured. Consequently,
the drilling of such shales, unknowingly under balanced, leads to instability problems. In
Figure 4R.8 we have an illustration of pore pressure and mud weight used vs. depth for a
well drilled offshore California from platform Hermosa. A detailed pore pressure analysis
indicates the traditionally troublesome shale interval on this platform between 5,000 and
6,000 feet to be abnormally pressured in the vicinity of 11.2 ppg pore pressure.
Traditionally this interval was drilled with 10.0 ppg mud or less with a great deal of shale
heavings and in some cases stuck pipe. A rate of penetration plot for this interval
appears in Figure 4R.9. The rate of penetration is on a logarithmic scale. Note that
penetration rates increase significantly and regress through this interval confirming the
presence of an abnormally pressured interval. Due to this analysis, a more recent well
was drilled with a higher mud weight closer to balancing the shale pore pressure. When
this was done, the shale heaving problem disappeared.
4. CONCLUSION
Borehole instability is a very expensive problem in our operations and throughout the
industry today. However, the majority of our problems appear to have a solution in an
increased understanding of the environment in which we are drilling and a better
awareness of surge pressures and pipe running speeds. The solution to many of our
instability problems can be solved through an increase in evaluation of the environment
we are about to drill and more planning. This can be developed through a detailed pore
pressure analysis, rock composition analysis via logs, mud logs, cores and paleo data,
and a better understanding of the structural relationship between wells surface to total
depth. Also a thorough analysis of allowable pipe running speeds to avoid surges as well
as a close eye on actual tripping speeds via real time data analysis is essential. This
would require a change in philosophy, as a company, from one of reactionary, and
rushing our drilling prospects from every planning aspect, to that of a detailed planning
approach and project team analysis.
Figures 4S.1 through 4S.5 summarize the type of information necessary to develop the
drilling model.
Figure 4S.1
Figure 4S.2
4. A bulk density log from at least one of the offset wells above. (In
addition, it would be desirable for the log to be integrated to
determine the relationship of Overburden Gradient vs. Depth and
have this information graphically displayed.)
6. Scout tickets, mud logs, any drilling summaries for the offset
wells above.
Figure 4S.3
9. Half scale (one inch = .4 sec), interpreted seismic lines which tie
the proposed location to the offset wells above.
Figure 4S.4
12. An ITT curve on at least one of the offset wells which see
abnormal pressure (pore pressures higher than the normal fluid
gradient), but preferably two offsets. The ITT on any offsets
should be displayed with the sonic logs of these offsets for
comparison. The vertical scale should be one inch = 1000 feet,
the horizontal scale should be two cycle, semilog, in micro-
seconds per foot.
13. An ITT curve at the proposed location with same scale as above.
NOTE: If well is directional, may need two or three ITT curves at
proposed location to cover well path.
Figure 4S.5
An additional pore pressure indication tool that can be used quite successfully is a
penetration rate plot or ROP for simplicity. Any time a mud logger is used on a well, they
plot some sort of ROP plot for correlation purposes. However, varying the scale can be
extremely beneficial or helpful in both correlating and in the determination of higher pore
pressures on a well. The illustration in Figure 4T.1 is a type of paper we use to plot a
particular ROP plot or a scale used. It is triple cycle semi-logarithmic paper and we plot
values of penetration rates from geolograph charts. Every 5 feet, we plot the average
penetration rate in minutes per foot along the semi-logarithmic scale, and we plot down
the depth scale or the vertical scale of 1 inch = 100 feet.
In Figure 4T.2, we have an SP curve on the left. The next two curves on the right are
resistivity curves and on the extreme right, we have replaced a conductivity curve with
such an ROP plot. Note the amount of character you see in this ROP plot in this
particular scale and how well it correlates with the electric log.
It is also useful to correlate ROP’s with offset ROP’s such as in Figure 4T.3, where we
have the ROP on a side track hole versus an original hole. Here we can also see the
significant changes in lithology which occur.
When we do determine pore pressures from logs, we have the capability of getting
reasonable accuracy as illustrated in Figure 4T.4. We have log derived pore pressures
and compare them to actual pore pressures as determined either from bottom hole
pressures or drill stem tests, and so forth.
You can see that we can get fairly accurate. Graphically illustrating this in Figure 4T.5,
you can see that log drive pore pressures and actual pore pressures can be fairly close.
In Figure 4T.7, we have an ROP plot. Notice at the bottom section, we have a gradual
increase in penetration rate drifting to the left. If we hold all things constant such as bit
weight, mud weight, rotary speed, etc., then there are only two things which can affect
penetration rate. One is a lithology change and the other is pore pressure. Lithological
changes from shales to sands are indicated by very rapid increases in penetration rate.
We pick up the sand tops in this fashion, however, the gradual increase in penetration
rate noted at the bottom of the hole, since this is a shale section, can only be due to an
increase in pore pressure. This gradual increase in penetration rate is not obvious when
plotted on other scales. If we plot on a linear scale, it is masked completely. This
particular scale exaggerates and picks it up quite well.
In Figure 4T.8, we have replaced a conductivity curve with an ROP plot from the original
hole and compared it to the log on the side track hole. Notice the resistivity on the side
track hole on the bottom is drifting to the left indicating higher pressure, and the ROP plot
is drifting to the left indicating that same high pore pressure interval in the original hole.
The reason the ROP plot on the original hole penetrates deeper than the side track hole
is no one was watching the ROP plot as we drilled off and lost the hole. It is one thing to
plot an ROP plot, but monitoring and analyzing the data as it is obtained, is necessary.
These increases in penetration rate are something to react to.
In Figure 4T.9, we have replaced the ROP plot with a conductivity curve and here again
we see the conductivity drifting to the left indicating higher pore pressure on bottom.
On another well, we have an ROP plot in Figure 4T.10. We have only plotted it down to
the top of the transition zone. Nothing appears to be happening so far on this well.
In Figure 4T.11, we look a little deeper on the electric log, and see a dramatic drift to the
left in resistivity and conductivity, illustrating a fairly significant transition zone. If we
replaced the conductivity curve with an ROP plot, we see the ROP drifting to the left
significantly in that shale section on bottom, indicating that pore pressures are increasing.
In using this curve, one needs to be aware of everything that is taking place on the well
as well as correlating to the offsets and keeping an eye on the lithology, as in Figure
4T.13. Here we have two sections of hole with gradual increases in penetration rate.
One starts at approximately 15,600 and another at 16,300 feet. They both appear to be
gradual increases to the left in penetration rate or possible transition zones.
As it turns out, the interval at 15,600 was a section of hole where gradual additions of
diesel to the mud from 0% to 4% took place. This increasing penetration rate was due to
the additions of diesel as used as a drilling lubricant. The interval at 16,300 is a
lithological effect as we see illustrated in Figure 4T.14. The character of the sand is a
stair step effect out to the left as noted on both the Gamma Ray and the Conductivity.
In the interval above 15,600 feet, we note is all shale. Repeating this data in Figure
4T.15, and replacing the conductivity with the ROP plot, we see that these apparent drill-
offs, or increases in penetration rate due to pressure, are not as they seem. One was
lithology, the other a reaction to changes in the mud system. Everything needs to be
taken into account in using the tool.
Another ROP plot appears in Figure 4T.16. Here at the bottom of this well we have a
gradual increase in penetration rate to the left below 11,200 ft.
And as we look at the electric log in Figure 4T.17, we see the resistivity and conductivity
drifting to the left as well.
The data is repeated in Figure 4T.18, where the conductivity has been replaced with the
ROP plot. Here again, we wee the increase in penetration rate coinciding with the drop
off in resistivity and conductivity.
In Figure 4U.1, we have an electric log and an ROP plot spliced in place of the
conductivity on the right. We have drawn solid trend lines down the left-hand side of the
resistivity curve and, at approximately 9250” we see a change in the general trend of this
curve. At this point where the trend line changes, indicating the fact that pore pressures
may be changing, we note on the ROP plot that the penetration rate slows down
significantly. We have somewhat of a hard spot with respect to the formations in the
vicinity above and below.
Generally, any time we see a change in the resistivity trend, it can indicate either the top
of pressure, or within an abnormally pressured section, a point where the rate of
pressure building has changed. For example, if pore pressures were building one-half
pound per gallon every 100 feet and then started to build one pound per gallon every 100
feet, this would be an inflection point or a point at which the trend would change as well
as a point where you would tend to see some hard or slow drilling. This is not to say that
all hard spots will mean a change in pressure, however, most changes in pressure will be
accompanied by a hard spot. This hard spot generally can only be recognized if using
this or some other exaggerated ROP scale.
In Figure 4U.2, we are deeper in the same well. We note a significant shift in the trend in
the vicinity of 10,100 feet, and as we look at the ROP plot, we see a significant hard or
slow spot.
Additional examples in the same well as we go deeper appear here in Figures 4U.3 and
4U.4, and again we see that as the trend changes or the rate at which pressure is
building changes, we have associated hard spots and this can be used as an indicator or
something to keep an eye on as we drill the well.
Additional information which both the conductivity log and such an ROP plot on this scale
helps to identify is relative permeability. Generally, in sand shale environments, the
better the break or kick to the left in each sand interval, the more permeable and porous
each sand tends to be. So, for example, in Figure 4V.1, the D-6A sand is much more
permeable than the D-6 sand since we have a much higher kick to the left in that sand.
More permability means greater ease to get stuck in that sand if differential sticking were
a problem. The same thing is illustrated in Figure 4V.2. From the ROP plot, the D-7C
and D-8 sands are more permeable than the D-7 since they have greater kicks to the left.
The D-9 sand gets back to a lower permeability than those above.
The same thing is illustrated with the conductivity log looking at Figure 4V.3, #D-17
sidetrack. The greater kicks on the conductivity log tend to be indicative of greater
permeability for those sand members.
This is not always true but the majority of the time when dealing with sand shale
environments, in wet sands, the conductivity curve, as well as the ROP plot will be a
permeability or porosity indicator. In the illustration in Figure 4V.4, note how well the
ROP plot takes the place of the conductivity curve in the previous Figure.
Note also that significant changes in lithology are picked up as in Figure 4V.5. On the
sidetrack hole, we have picked up a D-8 sand which was shaled-out down in the original
hole, and note some sands are more or less permeable than they were in the original
hole.
In Figure 4V.6, we have an ROP plot for a section of hole in abnormal pressure. The D-
10 sand is of 12 pound per gallon, the D-11 series 13 pound per gallon pore pressures.
We do not see any drill offs or increases in penetration rate above. This is due to the
fact that in knowing these pore pressures in advance, mud weights are brought up in
anticipation of these pore pressures in these formations. Consequently, no kicks or
wash-outs are taken or created. We, therefore, see a steady trend of penetration rate
vertically in the shale sections. If increasing mud weights were not required, we would
see a significant reduction in penetration rate. The fact that penetration rate holds
essentially constant is indicative of the need for the increase in mud weight.
It is important to be as specific as one can with these average penetration rates for each
5 foot interval to obtain the character necessary in the ROP plot. For instance, in Figure
4V.7, penetration rates for joints of drill pipe have been averaged out. We, therefore, get
a very blocky ROP plot which may make correlations much more difficult. It is better to
average out each specific 5 foot interval and develop as much character in the curve as
possible.
In Figures 4V.8, 4V.9 and 4V.10, we have a C-17 and C-17 sidetrack. With the electric
logs side by side in Figure 4V.10, we note how the lithology has changed from one
borehole to the next.
In the sidetrack well of Figure 4V.11, we see how one sand from 7400 feet to 7600 feet
has been masked quite a bit because of this full joint averaging of penetration rate. We
do not pick up the specific lenses that have the greater or lesser degrees of permeability
such as the sand down at 8100 feet where we did get more specific with the ROP plot.
In Figure 4V.12, at 9650 feet, pump troubles developed and it was necessary to cut back
on circulation rate. If we re-normalized the ROP after the cutback or shifted it over to
recalibrate it, then that ROP plot would look like Figure 4V.13 or a more drastic increase
in penetration rate. Hydraulics will, therefore, influence, as well as many other things, the
ROP plot. Holding things constants much as possible is a key to using it as an indicator.
In Figures 4V.14 and 4V.15, we have the electric log and the electric log with the ROP
plot spliced in place of the conductivity for the same well. If we know our pore pressures
and we have them nailed down for each formation, then we can drill our wells in a fashion
where we prepare a weight up schedule and see to it that we have the proper mud
weight to balance each sand as we enter it. Granted, we have the capabilities of
handling kicks and desire to drill with as low a mud weight as possible. However, if we
know the pore pressures, then we accomplish drilling with as low a mud weight as
possible by staying correlated and weighting up accordingly. One of the advantages of
drilling into a permeable formation at least balance or slightly over-balanced is in the
avoidance of kicks.
While circulating out many kicks, we tend to get stuck. Also, as we drill the shale above
the sands under-balanced, we cause/create washouts which ultimately affect the cement
job, assist communication between productive sands and wet sands, and so forth.
Consequently, we can minimize hole washout and other problems if we can drill into each
formation either balanced or just a hair over-balanced. In hard rock, course, we chose to
drill under-balanced until permeable formations are encountered.
Figure 4V.14
As an example of how problems can arise, in Figure 4V.16, we have an ROP plot for a
well where we drilled into a formation with an 11 or ppg mud and took an 11.4 ppg
equivalent quick, ECD was not sufficient to control BHP, so it was, therefore, necessary
to circulate out on the choke. During the course of circulating out of the kick, trouble
developed with both rig pumps, one pump on the cementing unit, and two jets on the bit
plugged. This resulted in 36 hours to circulate out this kick. Many problems can arise at
the wrong time. Even if a kick is easily circulated out, there is a disadvantage in the
washouts created, and the wasted costly rig time.
In Figure 4V.17, we have an electric log on this well. Figure 4V.18 illustrates the ROP
plot, spliced in place of the conductivity curve. During the course of drilling, we want to
be verifying the pore pressures whenever we log at each casing point.
This can aid in the avoidance of many other problems, as well. For instance, in Figure
4V.19, we have a productive sand at 5250 feet. This sand was drilled with a 9.5 pound
per gallon mud. It is natural to assume being as shallow as it is, and being drilled with a
9.5 ppg mud, it to be a normally pressured or a nine pound per gallon equivalent pore
pressure. However, in plotting the logs, it turns out, this sand is a 9.5 pound per gallon
pore pressure.
In testing the well, the sand was perforated with a wire line casing gun with a 9.5 ppg
clear brine in the hole. Due to temperature expansion of the brine, the effective weight at
TD was less than 9.5 pounds per gallon. All brines do experience thermal expansion and
ultimately will have a reduced hydrostatic head due to this expansion. It would,
therefore, be necessary to know that this formation is of 9.5 pound per gallon pore
pressure and realize that something greater than 9.5 ppg would be necessary.