Practical Design of a Continuous Distillation Plant

for the Separation of Essential Oils from Aromatic Raw

Materials.

Table of contents

I. Introduction

II. Process theory and operational data

III. Practical design calculations and equipment considerations

IV. Cost estimates for various plant capacities

V. Conclusion

by: G. R. Boucard and R.W. Serth

Texarome Inc.

Leakey, Texas

e-mail address: texarome@hctc.net

internet address: www.texarome.com

Summary. A General Mass Transfer Cycle (GMTC) proposed by G.R. Boucard and R.W.
Serth has been adapted to the continuous steam distillation of essential oils from aromatic
plants. The process consists of a totally insulated pneumatic conveying system using
superheated steam as a carrier gas, and arranged in such a way as to provide a two stage,
counter-current flow of the gas and the solid phase. During the transport, which is made to
last 30 seconds, the oil goes into the vapor phase and exits the system with the steam after
filtration. Following total condensation of the gas stream, the oil is separated from the water
condensate in a gravity separator. The water phase is treated and recycled to the boiler. The
dry pulverized stream of spent solids is blown directly into the boiler biomass furnace to
generate steam and electric power for the process. All the components and the design
characteristics of the system follow the laws of pneumatic conveying, which affords a quick
and convenient way of scaling the system up and down. Shortcut calculations and cost
estimates are provided for the construction of plants with throughputs of 200 tons/day down
to 6.25 tons/day.

I. Introduction

The general concept of this technology and its application to the distillation of Texas
Cedarwood oil was described in an earlier paper by G.R. Boucard, President of Texarome Inc.
and Dr. R.W.Serth of Texas A&M University - Kingsville1 . At that time, care was taken not to
divulge the design and construction details of the continuous apparatus, which was in the
development stage. Hence, the proprietary process was successfully guarded from competitors
despite the absence of any patents or other forms of copyright protection. Since then, the
technology has generated considerable interest from distillers all over the world, but was
never released for licensing.

With the recent advent of the internet and an apparent worldwide demand for high value
natural extracts from various botanicals, not the least of them being essential oils, Texarome's
web page at : www.texarome.com , has generated a new flood of global inquiries from
distillers and academics alike. Apparently, there continues to be a lack of information on both
conventional and innovative technology for the distillation of essential oils. We hope that
releasing more details on the concept and design of the Texarome distillation technology and
its underlying theory will be of some benefit to other investigators seeking to improve the art.

II. Process theory and operational data

Entrepreneurs and businessmen are often bored by too much scientific theory, and just want to
build a plant that works, to make a needed product, marketable at a high profit. Academics, on
the other hand, can feed and ruminate on theory until the end of time while loving every
minute of it, and they are indispensable in testing the validity of alleged innovations, as well
as finding other applications, including avenues of possible improvement. Nevertheless, this
paper is primarily intended for entrepreneurs and potential distillers looking for practical
information on the construction of a modern essential oil distillation plant.
As beautiful as "high-tech" can be in the world of electronics, aviation, and medicine, "low-
tech" still has many advantages in very large scale raw material processing, especially under
field conditions. For instance, much praise has been given to the elegant technology of
supercritical CO2 extraction for delicate essential oils, and duly so, but steam is hard to beat
as an extraction fluid. A good old steam boiler and the plumbing that goes with it is
demonstrably more "user friendly" and less expensive than supercritical CO2 compressors and
the intricate piping and sealing arrangements required to and from the high pressure CO2
batch stills.

The basic concept of Texarome's continuous essential oil distillation process evolved from the
following considerations, which were based on prior laboratory research by the authors 2 .

Step 1. If one can find a way to put finely ground aromatic material (any essential oil
plant tissue) in continuous contact with low pressure (2 psig) superheated steam, and
maintain the contact for 30 seconds in a well dispersed two phase system, the oil will
vaporize and travel from the plant tissue (solid phase) into the steam environment (gas
phase).

Step 2. If one can find a way to perform step 1 twice, or to carry out the contacting in a
counter-current fashion, then the plant tissue will be virtually exhausted of its oil
content, and a maximum yield of oil will be obtained.

Step 3. Having achieved step 1 and step 2, it is now necessary to find a way to end the
contacting and to cleanly and totally separate the gas phase (containing the steam and
all the oil) from the solid phase (exhausted plant tissue of any essential-oil bearing
plant).

Step 4. This is a conventional step which consists of piping the filtered gas phase (oil and
steam) into any type of water-cooled or air-cooled condenser, to obtain a mixture of oil
and water condensate, which then can be separated in a gravity decanter, also known as
a Florentine flask.

It sounds simple and it is simple. Indeed, it is quite possible that other continuous contactors
and extractors, already available on the market for other tasks, may well satisfy the above
conditions, with some minor adaptation. We found, however, that Texarome's apparatus was
very inexpensive to build. We should add that the process described herein is just the "drying"
and "steam distillation" application of what we like to refer to as a General Mass Transfer
Cycle (GMTC). As can be seen in the drawing below (Figure 1), the configuration of the
general apparatus lends itself to a number of continuous mass transfer tasks, such as:
multicomponent distillation (vapor/liquid), gas absorption (gas/solid, gas/liquid), leaching
(liquid/solid), drying (gas/liquid), extraction (liquid/liquid), partial pressure distillation
(gas/liquid and gas/liquid/solid), evaporation (gas/liquid) and even crystallization, desorption
and ion-exchange.
A. The General Mass Transfer Cycle (GMTC) and the explanation of Figure 1

The Boucard/Serth concept of a general mass transfer cycle is illustrated here on a distillation
column to explain how Texarome's continuous process for the steam distillation of essential
oils is related to this general idea.

The mass transfer application probably most familiar to chemical engineers is multi-
component distillation, more specifically, the separation of petrochemical mixtures or the
separation of various other mixtures of volatile organic compounds (VOC). Textbooks such as
Unit Operations of Chemical Engineering3 deal at length with the theory of mass transfer and
with the various computational methods used in distillation operations. The standard
nomenclature of column calculations is used in this example to help situate the parameters of
the process and identify the hardware components of the apparatus. With reference to Figure
1:
 A complete STAGE is circled with a curved line.

A TRAY is shown within a dotted line.

Also shown are the REBOILER, the CONDENSER, and the REFLUX.

In this case, the two phases are the LIQUID and VAPOR PHASES (L and
V).

The FEED point is shown for continuous operation.

Finally the DISTILLATE and the BOTTOMS are shown as the two end
products.

The distinction of this GMTC concept is that, using cyclonic separators or hydroclones,
it accommodates all phase combinations (vapor/liquid, gas/liquid, liquid/liquid,
solid/liquid, gas/solid) and is, at least in theory, applicable to most mass transfer tasks,
wherefore the acronym GMTC for General Mass Transfer Cycle.

The design superiority of a mass transfer apparatus with such a positive conveying
configuration of "kinetic fluids", moving co-currently within a stage, and counter-currently
between the stages is seen as follows:

1) Centrifugal separation of L and V phases within a stage, using a cyclonic vessel (a),
assures no flooding or foaming, and minimal entrainment.

2) Venturi- type atomizing nozzles (c) provide excellent dispersion of L and V phases,
equivalent to nearly 100% tray efficiency.

3) Rotary valves or sealed magnetically driven gear pumps (c) replace the downcomer
pipe and deliver the liquid of a stage into a venturi device (b) to be "atomized" by
pressurized vapor from the stage below.

4) The high velocity of the fluids assures less fouling of all the contact surfaces, and the
negative implications of fouling are diminished.

5) Better control of all intensive variables, because of positive (kinetic) displacement of all
fluids, compared with a gravity driven system. Very high pressures are still economical for
large flow rates.

6) Simple horizontal construction on skids for quick and easy installation; less crane time.

7) Modular design of stages allows easy design error correction or process modifications.

8) Dismantling of system allows re-use of modular components for other processes. Even
piping connections are identical, or increase in exact increments.

9) Modular tanks are easy to fabricate and mass produce for quick assembly in the field, in
three or five standard sizes.
10) Easy rectangular box-in insulation (plenum) for the entire apparatus with immediate (on-
the-ground) door access for repairs, maintenance and cleaning.

11) Considerably less expensive than column at equal capacity.

As with any technology, there are some disadvantages to this mass transfer apparatus, and
these are:

1) The necessary addition of a moving part: the rotary valve or the gear pump and its
mechanical drive.

2) The higher pressure drop between stages, which complicates calculations, may increase
operating costs, and makes the device unsuitable for vacuum separations.

3) The apparatus becomes cumbersome and unwieldy for separations requiring a very large
number of stages.

Figure 2 shows a three-stage GMTC for any gas/solid system.

B. The Texarome continuous steam distillation process as a proven application of the
GMTC concept

Entrepreneurs and distillers may skip the above section II.A., as it is only intended as an aside
for chemical engineers, who may wish to test Texarome's proposed GMTC against rigorous
methods of calculation and modeling for the different separation tasks of interest to them. In
the meantime, non-scientist, essential oil distillers can review Texarome's specific application
of the GMTC concept to steam distillation, and thereupon evaluate the possibility of
retrofitting their plants from a batch to a simple continuous process, or building new
continuous plants.

The physical laws underlying the process itself fall obviously under the domain of chemical
engineering, but the actual apparatus obeys the laws of mechanical engineering, in particular
those governing pneumatic conveying. If the distiller is already familiar with pneumatic
conveying this will be easy to understand. If not, consider this brief definition:

A sufficiently strong flow of air (or any dry gas) generated by a compressor or a blower,
is able to pick up a certain amount of finely ground solids (or liquid droplets) and blow
them through a pipe from point A to point B over long distances. This is generally called
pneumatic conveying (Greek, pneuma, breath).

Texarome's process is basically a pneumatic conveying system which uses superheated steam
instead of air as a carrier gas. We found that steam conveys just as well as air, provided it is
not allowed to condense. One of the key requirements of this system is that it must be
totally and thoroughly insulated. Since steam coming from a boiler is already under
considerable pressure, no blower or compressor is needed. In fact, a pressure regulator must
be used to bring the steam pressure down to 2 psig, a standard pressure in pneumatic
conveying. An important attribute of steam is that it can be easily superheated, which makes it
"extra hot". Superheated steam contains "sensible heat" and "latent heat", both of which
together are known as the steam enthalpy. Now, "sensible heat" is "extra heat", so to speak. It
is the part of the steam enthalpy which can be transferred to (cold) surrounding particles to
vaporize the oil in them by virtue of the partial pressure effect (Dalton's law) without the
steam condensing back to water. If

one starts drawing on the "latent heat" of the steam to heat up the surrounding particles and
the apparatus, the steam will condense on the particles and the pipe wall, and the conveying
will run out of "pneuma", resulting in a plugged apparatus . So the objective is to calculate a
mass ratio of superheated steam and distillation material (conveyed material) for which the
sensible heat portion of the steam will suffice to heat up the incoming distillation material
rapidly to 212 F and eventually to the system's temperature of , say, 400-425 F, without
condensing. It is a simple calculation if one takes the specific heat of dry biomass to be
generally 0.5 Btu/lb/degF. Calculations and practice have shown us that a ratio of 1: 1 (one
pound of steam/hr for one pound of dry [10%] distillation material/hr) not only provides
enough heat to vaporize the oil, but it so happens that this amount of superheated steam at 425
F and 1 to 2 psig constitutes enough gas volume, ( the steam tables4 show that one pound of
superheated steam at 425F and 2 psig occupies a volume of approximately 32 cubic feet) to
convey the material through the system pneumatically at the standard conveying velocity of
ca. 5000 fpm, given the appropriate size of pipes. It must be noted here that ground up
materials with a moisture content higher than 10% and up to 50% can be processed just as
well, so long as they remain a free flowing solid, with an angle of repose of 60 degrees or
less. The only difference is that the wetter material may require a higher ratio of superheated
steam, since the higher moisture content will have a tendency to quench (de-superheat) the
steam and lower the system's temperature below ideal conditions for partial pressure steam
stripping. Nevertheless, the ratio is likely to be lower than in conventional batch distillation,
which for purposes of comparison requires ratios of 3:1 for cedar, 12:1 for vetiver and 20:1
for amyris.

The Texarome apparatus, just like any pneumatic conveying system, consists of the following
components:
1) a blower (actually a boiler in this case)

2) conveying pipes

3) cyclones

4) rotary valves (airlocks)

5) a dust filter

However, aside from the special way of running the conveying pipes, there is one more key
element in Texarome's apparatus which must be added to the conventional pneumatic
conveying system to make it distill essential oils, and that is:

6) residence time cyclones

Chemists know that all reactions take time to complete, from very short reaction times like
explosions to long reaction times like fermentation. But physical separations also take time,
from a "flash" drying to a week-long sedimentation. Texarome's continuous distillation is a
"flash drying" or "flash distillation" of sorts. However, laboratory tests conducted under
different operating conditions showed that the flash occurs in a steep curve which, at a
temperature of 400 F, plateaus after a period of 30 seconds, given a particle size of 35 mesh
and under (for wood)2. Distillation materials with the oil on the surface of the plant tissue,
such as certain leaves and flowers, may require shorter "flash times". At any rate, a standard
pneumatic conveying system, designed for the purpose of conveying only, does not provide
sufficient residence time for the "flash distillation" to take place in most cases, but certainly
not in the case of cedarwood. For that reason, a practical way had to be found for "making
time" inside the system.

In this case, two intensive variables affect the required residence time for "flash distillation":
temperature and particle size.

Temperatures higher than 450 F will cause charring and partial pyrolysis of the biomass and
the decomposition of the essential oil. Temperatures of less than 400 F will significantly
lengthen the required residence time for equal yield. Hence, 400 to 425 F was found to be the
ideal operating temperature for flash essential oil distillation. (In applications where this
process would be used to steam strip undesirable organic volatiles such as toxic substances
from contaminated soils for instance, superheated steam up to 1000 F could be used for a
thorough stripping, and a decomposition of the toxic compounds may even be welcome.)

Particle sizes larger than 35 mesh will drastically increase the required residence time.
Particle sizes smaller than 35 mesh will shorten the residence time and increase the yield at
equal temperature, but are very hard to achieve economically with heavily lignified biomass
such as woods. Dry leaves, seeds and roots are even better candidates for this system than
wood because they pulverize easily.

Thus, 30 seconds has been found to be a good distillation time in Texarome's continuous
process. But how does one achieve this much residence time in a system where the material
travels at 5000 fpm (83 feet per second), and still make a 100 ton/day plant fit on a 40 foot
trailer? The authors had made the earlier observation that material entering a cyclone spends a
significant amount of time within the device. In fact, some material stays spinning around
inside the cyclone even after cutting off the feed of in-going material, due to the updraft
created by the vortex, and only falls out after the air flow from the blower is stopped.

Using a standard high efficiency cyclone design with plexiglass windows, the authors studied
the flow pattern and residence time of various particle sizes in a cyclone.It was established by
direct measurement that biomass solids of 35 mesh and under with a moisure content of 10%
and a density of 15 to 20 lbs/cuft entering a standard high efficiency cyclone (see Figure 4) at
the standard conveying velocity of 5000 fpm will reside in the cylcone for at least 0.5 seconds
for each linear foot of cyclone height.

Hence, one can see that, whereas a linear foot of standard pneumatic conveying pipe in a
system of any size provides only 0.072 seconds of residence time, a small 6ft cyclonic vessel
provides 3 seconds, and a larger 12 foot cyclone provides 6 seconds. Obviously, the larger the
system, the smaller the number of residence time cyclones required. Or, put in business terms,
it pays to go big. Texarome's 1000lbs/hr apparatus has a total of 9 cyclones 6 feet tall. Larger
systems may require a total of only 4 or 5 cyclones. By comparison, it would take half a mile
of pipe to achieve a 30 second residence time.

The above explains the need for the 6th pneumatic conveying component introduced by
Texarome: the residence time cyclone.

III. Practical design calculations and equipment considerations

Texarome's apparatus consists simply of a series of pipes connected to small cyclonic vessels
and three rotary valves ( Figure 3 ). Pipes and empty vessels are inexpensive. Rotary valves
are relatively expensive unless they are built by a local machinist. The novelty of the design is
a rather convoluted way of running the pipes and interconnecting the vessels in order to
achieve a two stage, counter-current flow of the phases (gas/solid), with subsequent
separation of the phases, while maintaining a co-current flow for solids transport.
Furthermore, the residence time cyclones are used to provide the contacting time needed for
the mass transfer to take place. Although our apparatus faithfully incorporates the three main
requirements of the multistage mass transfer process (counter-current flow, phase contacting,
and phase separation), and is a serious piece of chemical processing hardware, any small
welding shop can fabricate and assemble such an apparatus using carbon steel or stainless
steel, and deliver an essential oil distillation plant of considerable throughput in a very short
time, for a fraction of the cost of a conventional batch plant with the same capacity.

The typical e-mail addressed to Texarome will read: "we have a lot of caraway in Holland", "a
lot of eucalyptus and tea tree in Australia", "a lot of lemongrass in India", "lot of vetiver in
Haiti", "a lot of coriander in Russia", "a lot of bay leaf in Puerto Rico", "a lot of cedar leaf in
Canada", "a lot of red cedar in Oklahoma", "what will it cost me to build a continuous plant?"
Understandably, the question also implies: how much money am I going to make? Of course,
these questions cannot be answered for all the tea in China unless the entrepreneur gives the
following specifications:

1) What is the desired capacity of the plant, expressed in tons/day of raw material?

2) What is the bulk density and the moisture content of the raw material?

3) What is the approximate content of essential oil in the raw material in weight %?

Other information of relevance for the plant's profitability is:

4) What are the boiler fuels available and at what price?

5) What is the cost of electricity?

6) What is the cost of skilled and unskilled labor?

7) What are the regulatory requirements for air and water emissions?

But really, the basic plant can be designed with specifications 1) and 2). Let's take the case of
an existing producer, possibly contemplating to retrofit his batch plant to process 200 tons per
day of cedarwood with Texarome's continuous distillation system and run it with only 2
operators per shift, making 30% more oil than he now makes at equal throughput. Cedarwood
is one of the toughest material to flash distill because of the high vapor pressure of the oil, its
virtual encapsulation in the lignified cell tissue of the wood, and the high cost of grinding.
This is to say that this design example will encompass the

processing of many other essential oil bearing raw materials in addition to cedarwood.

Since the continuous plant runs around the clock in a fully automated mode, let's take a
nominal throughput of 200 tons/day divided by 24 hours, which is 8.3 tons/hour or 16,666
lbs/hour. At a yield of 2.5% , such a plant would produce 417 lbs of cedarwood oil per hour,
10,000 lbs/day, or 22.7 drums/day ( 55 gal drums/440 lbs net). From this 200 tons/day design
example, we will provide a table below to show dimensions for scaled down plants of 100
tons/day, 50tons/day, 25 tons/day, 12.5 tons/day, and 6.25 tons/day. A flow sheet for a
complete plant layout, of any size, is shown in Figure 5.

1) Sizing the feed bin

Any continuous system must be fed evenly from a holding bin or feed bin. Such feed bins are
usually fitted with a variable speed drag-chain or a feeder screw which meters the same
amount of material into the continuous system around the clock. In this case, the feed rate is
16,666 lbs per hour. Bins with large capacities are very expensive and troublesome. It is best
to size a bin to hold only 1 to 2 hours of feed and replenish it with a front-end loader from a
large pile of pre-ground raw material located inside a building. Hence, this project calls for a
16 ton or 65 cubic yard bin. However, the bin should hold the coarser material from a first
grinding, and then meter the coarse material into two 250 hp hammermills for a second fine
grinding (35 mesh and under), which then feed the continuous system through a rotary valve.
Bins are just rectangular steel boxes with V-shaped bottoms and vibrating sides, quite
simple and inexpensive to build, except for the chain or auger drive.

2) Sizing the steam generator (boiler)

Given the established steam/raw material ratio of 1:1, the process will require 16,666 pounds
of steam per hour for 16,666 lbs of distillation raw material per hour. Since one boiler
horsepower delivers roughly 34 lbs of steam per hour, the boiler size required is 16,666
divided by 34, or 488 boiler horsepower. Thus, a 600 hp boiler will be adequate. The boiler
must be rated at 250 psig operating pressure and be fitted with a superheater coil capable of
reaching a steam temperature of 500 F. (To scale down the system's boiler to any size plant,
use the 1:1 ratio.) Unless the spent distillation material has a higher market value than local
fuels (by Btu comparison), the boiler can be fitted with a biomass furnace which will use all
the dry pulverized distillation material as "free" fuel to generate steam for the process and
electric power for the entire plant. A 400 kW Turbine Generator Set (TG set) can be used as a
first stage pressure reducing device, and the turbine exhaust (extraction steam) at 15 psig
further reduced to 2 psig for the process. However, this design example foresees only the
biomass furnace, which adds roughly 50% to the total boiler cost, as compared with a gas/oil
fired boiler. The TG-set and switch gear can be added to the plant for about the same cost
(used) as that for the total boiler installation.

3) Sizing the pneumatic conveying components

a) Pipe size. In order to establish the size of the conveying pipe, one must determine how
16,666 lbs/hour (278 lbs/minute) of superheated steam at 425 F and 2 psig translates into
cubic feet per minute of conveying gas, moving at 5000 fpm. For that, we go to the steam
tables and find, as previously noted, that one pound of superheated steam at 425 F and 2 psig
has a volume of approximately 32 cubic feet. Hence, the flow rate of the conveying gas is
278 x 32, which equals 8896 cfm and requires a pipe of 18 inches in diameter to obtain a
proper conveying velocity of 5000 fpm. (One linear foot of an 18 inch pipe has a volume of
1.76 cuft, and if 8896 cubic feet have to be pushed through the pipe in one minute, the gas
velocity will be 8896 divided by 1.76, which is equal to a conveying velocity of 5054 fpm.) A
scaled drawing will reveal exactly the total length of pipe required to connect the cyclones
together. As discussed earlier, 30 seconds residence time means a total of 60 linear feet of
cyclone height.

b) Cyclones. Consistent with the above pipe size calculation, the recommended inlet pipe
diameter of a standard cyclone rated for 8,150 to 10,000 cfm (see Figure 4) is also 18 inches.
Actually, given the steam volumetric flow rate (in cfm) the cyclone inlet dimension listed in
the cyclone table (Figure 4) can be used to size the pipe, and to scale the whole system up and
down. This brings us to the other cyclone dimensions as shown in Figure 4. The required
cyclones for 8150 to 10,000 cfm have a diameter of 80 inches and an overall height of 21
feet. We know that this means a residence time of 10 seconds each. Thus only 3 cyclones are
required; more precisely, two residence time cyclones and one cyclone separator (discharge
cyclone). A third residence time cyclone could be added (making a total of four cyclones) to
provide a service factor of 1.3, but for this example we will use a total of just three cyclones.
To estimate the price of a cyclone, take the height and diameter of the cyclone, and figure the
area of a cylinder of the same height and diameter. This will determine approximately the
weight of steel required, which can then be multiplied by the price of steel (or stainless steel).
Add $1.00/lb for fabrication.
c) Filter. The filter for this plant requires a filtering area capable of handling the gas (steam)
flow rate of 8896 cfm. Our operating experience shows that one standard 6 inch diameter bag,
8 feet long can handle 40 cfm, being pulsed every five minutes for 0.5 seconds. By the way,
this is nearly twice the area required for standard air filtration, due to the inevitable caking of
the wet particles on the filter cloth during repeated cold start-ups. A set of bags will last
approximately six months. The bag material must withstand temperatures of up to 450 F.
Hence, the flow rate of 8896 cfm will require (8896 : 40 = 222) a steam jacketed filter
cyclone with 222 high temperature bags. The baghouse tube sheet can accommodate 1.52
bags per square foot, or 1 bag per 0.66 sqft. This determines the diameter of the filter cyclone
for cost estimation. One must bear in mind that the filter cyclone has to be entirely steam
jacketed, which for cost estimation purposes is equivalent to two cyclones.

d) Rotary valves. The rotary valves required for this system are valves with a volumetric
displacement which, at 10 rpm (a good airlock speed), is equal to three times the volume of
material throughput (valve pockets must run only one third full to prevent plugging). Since
we are running 16,666 lbs/hr or 278lbs/min, and since the average density for cedar is about
18 lbs per cubic foot, we divide 278 by 18 and obtain a volumetric throughput of 15.4 cubic
feet per minute of ground-up cedar. The rotary valve must be rated 46 cfm at 10 rpm, or 4.6
cubic feet per revolution. This is approximately the capacity of an 18 inch valve, available
from a variety of vendors. However, a rule of thumb is to size the valve for the discharge
diameter of a standard cyclone, which in this case is 20 inches. Thus, three 20 inch rotary
valves are required. The rotary valves are the only moving parts of the entire continuous
apparatus, but they are also a key component which must be properly selected from a variety
of designs and operate perfectly. Without a good seal provided by the rotary valves at the feed
inlet and at the discharge, the system will not operate efficiently.

4) Sizing the cooling system

The cooling system must dissipate roughly 1100 Btu per pound of steam exiting the system at
approximately 350 F, that is, 18,300,000 Btu/hr. Using a shell-and-tube heat exchanger,
approximately 1,000 sqft of heating surface is required to condense the steam flow rate of
16,666lbs/hr. A 75 hp pump is required to circulate the water from the cooling tower (or spray
pond) to the heat exchanger. The cooling tower is a simple device for the distiller to build
himself, either with a natural draft or a forced draft using a fan. But it can also be just a spray
pond of 100ft x 200 ft with 200 sprinkler heads of 5 gpm each. The problem with open
cooling towers or ponds in a dusty environment of grinding and material handling, is the
contamination of the cooling water with solids, algae, etc., which tends to foul the shell side
of the heat exchanger. There is also the necessity to make up for evaporation with treated
water, which is expensive. If at all possible, large radiator-type air-cooled heat exchangers
should be used to re-circulate the cooling water with virtually no fouling and no evaporation.
Based on experience and practice, our shortcut calculation for sizing cooling
requirements is that each 1 gpm (500lbs/hr) of condensate calls for 30 sqft of heat
exchanger surface, 2 hp of pumping power, or 30 gpm of 100 F cooling water at 45 psi.

5) Sizing the oil separators

This is a conventional unit operation unrelated to Texarome's continuous process, but it merits
some attention because this is where the money (oil) is collected. At a steam flow rate of
16,666 lbs per hour, the condensate flow rate will be 16,666lbs/hr, or 33.3 gallons per minute.
If the separator (Florentine type) diameter is too small, the downward velocity of the water
column in the tank will exceed the upward velocity of the oil droplets, and a good part of the
oil will be entrained and lost. Very light oils, such as citrus oils, separate readily at ambient
temperatures in rather small receivers, but we have found by observation and empirical
methods that oils with a specific gravity larger than 0.98 require a maximum downward
velocity of the receiver water column of 0.25 inches per minute, and a receiver temperature of
180 to 190 F. Of equal importance is the proper distribution of the distillate inside the
receiver, so as to avoid currents and eddies. The installation of a coalescing element will also
improve separation of the oil and water phases.

IV. Cost estimates for various plant capacities

There are several ways for a prospective entrepreneur to obtain a cost estimate on a plant. The
easiest and perhaps most elegant way is to pick up the phone and call a large international
engineering firm such as Bechtel Corporation or Brown & Root, submit some sketches and
specifications, and request a quotation for a turnkey plant. This is generally the most
expensive route to a finished plant. At the opposite end of the spectrum, the least expensive
way is to set up a small fabrication shop and start collecting equipment and materials
purchased in cash from used equipment dealers in order to build a plant that is just as pretty
and efficient as one built by a large engineering firm. This is known in Texas as the "poor
boy" way. There is no cheaper way, provided the entrepreneur has the necessary engineering
talent, the ingenuity and the working disposition to venture on that path. Another alternative
for large and midsize companies is to subcontract the fabrication of the various components
and sub-components, and have in-house engineers assemble them, once they arrive on site.
But as a rule, a used boiler with clean tubes is just as good as a brand new one. A used electric
motor with good windings and new bearings is as good as a new one. A used hammermill in
good condition with new hammers, new bearings, and new screens is as good as a new one.
All of the above can be obtained for 25-50% of new price. The same applies to used pipes,
whether carbon steel or stainless. On the other hand, structural steel and sheet metal are so
relatively cheap that it is best to buy them new and immediately apply metal primer to the
surface prior to fabrication, which allows outside storage pending use in the fabrication shop.

Equipment costs for continuous essential oil distillation plants of various capacities are
summarized in Table 1. These cost estimates are based on the design calculations given above
and the authors’ 25 years of experience in plant fabrication. They are meant to serve as a
guideline for distillers wishing to have a plant built. All equipment is quoted used and in good
condition. Used equipment prices are listed at roughly 50% of new. The fabricated
components, such as the cyclones, feed bin, oil separators, the filter housing, etc., are quoted
at $1.30/lb for carbon steel and $2.75/lb for stainless to cover material and labor. The airlocks,
filters and filter cages are quoted new.

Another purpose of Table 1 is to give an idea of what it should cost to have a small local
engineering construction firm handle the turnkey fabrication of these plants. Typically,
doubling the raw cost of material, labor, and shop overhead, as listed in the table, should
constitute a fair turnkey price, either for the whole plant or the components. However, this
may not include the start-up and "de-bugging" of the plant, since the builders are not likely to
be familiar with this particular technology, or even with essential oil distillation in general.
Thus, for our example plant processing 200 tons/day of cedarwood, the total turnkey cost
would be about $830,000.

Finally, it is customary to expect to pay a premium over and above the fabrication price for
proprietary technology. This premium is usually factored into the total project in the form of
up-front cash, royalties, licensing fees, stock ownership, or other form of payment for the
transfer of the technology from the developer to the end user.

ITEM 200 100 tons/day 50 tons/day 25 tons/day 12.5 tons/day

tons/day
8,333 #/hr feed 4,166.5#/hr feed 2083 #/hr feed 1041 # /hr feed
16,666 #/hr
feed 462 cu ft/ h feed 231 cu ft/ h feed 115 cu ft/ h feed 57 cu ft/ h feed

925 cu ft/ h 4444= cfm steam 2222 = cfm steam 1111 cfm steam 500 cfm steam
feed
 8888 = cfm 16.7 8.3 4.1 2 gpm dist.
steam gpm gpm gpm
dist. dist. dist.
33.4 = gpm
dist.
METERING BIN 1850 cuft 925 cuft 462 cuft 231 cuft 115 cuft
AND
8’x8’x29’ 8’x8’x15’ 8’x8’x8’ 8’x4’x8’ 8’x4’x8’
HAMMERMILL
2 X 250 HP 250 HP 125 HP 75 HP 40 HP
COST $100,000 $50,000 $35,000 $25,000 $20,000
BOILER AND 490 HP 245 HP 122.5 61.2 30.6
PIPING HP HP HP
600 HP 300 HP nominal
nominal 150 HP nominal 80 HP nominal 40 HP nominal
COST $60,000 $40,000 $30,000 $20,000 $15,000
SS 120 ft of 18" 120 ft of 12" pipe 120 ft of 10" pipe 120 ft of 6" pipe 120 ft of 4" pipe
PIPES/ELBOWS/ pipe
17 elbows/flanges 23 elbow/flanges 29 elbows/flanges 38 elbows/flanges
FLANGES 14
elbows/flanges
COST $25,000 $20,000 $15,000 $10,000 $5,000
SS CYCLONES 3 each - 80" 4 each - 54" dia. 6 each - 42" dia. 8 each - 28" dia. 11 each - 20" dia.
PLENUM dia.
14 ft height 11 ft height 7’6" height 5’4" height
21 ft. height
COST $55,000 $50,000 $45,000 $35,000 $25,000
FILTERS 222 bags 111 bags 56 bags 28 bags 14 bags

AND SS 146 sqft tube sheet 73 sqft tube sheet 36.5 sqft tube sheet 18.25 sqft tube sheet 9 sqft tube sheet
HOUSING
COST $35,000 $25,000 $16,000 $8,000 $4,000
ROTARY 18" 14" 12" 10" 8"
VALVES
COST $60,000 $45,000 $35,000 $25,000 $15,000
COOLING PUMP 75 HP 1000 40 HP - 480 gpm 20 HP 240 gpm 10 HP 120 gpm 5 HP 60 gpm
hp gpm
10,000 sqft pond air cooled air cooled air cooled
AND 20,000 sqft
TOWER/POND pond
COST $40,000 $20,000 $15,000 $10,000 $5,000
SS CONDENSER 1000 sq ft 500 sq ft 250 sq ft 125 sq ft 62.5 sq ft
SURFACE
COST $5,000 $4,000 $3,000 $2,000 $1,000
SS OIL 4 X 9.25 ft dia. 2 x 9.25 ft dia. 9.25 ft dia. 6.6 ft dia. 4.8 ft dia.
RECEIVER
COST $20,000 $10,000 $5,000 $2,000 $1,000
ELECTRIC $50,000 $30,000 $20,000 $10,000 $5,000
SYSTEM
 TOTAL $415,000 $294,000 $219,000 $147,000 $96,000

V. Conclusion

Although the information contained in this paper may be of some use to experienced
construction engineers and fabricators, we must caution all others that there is a considerable
amount of special know-how and important construction details, too numerous to address
here, which will make such a processing plant work right away or work much, much later. Be
aware that it took Texarome's experienced distillers years to totally perfect this continuous
distillation system. Finally, for the benefit of those outside the field of essential oil distillation,
it should be mentioned that this process, which also falls under the broad classification of
"thermal desorption" processes, is immediately adaptable to stripping any volatile organic or
inorganic compound with a boiling point of up to 1000 F from any solid matrix suitable for
pneumatic conveying. For that reason, it has been looked into by the U. S. Environmental
Protection Agency as a possible method of treating contaminated soils, and is listed on the
Superfund's VISITT data base as an innovative treatment technology.

For all legal purposes, Texarome, Inc. is the developer of this innovative technology for
the continuous distillation of essential oils and for other mass transfer tasks. Interested
parties are invited to discuss the terms of a technology transfer for any commercial
application of the process.

References

1. G.R. Boucard and R.W. Serth, Perf & Flavorist, 16(2), 1-8 (1991).

2. R.W. Serth, G.R. Boucard and B.S. Ainsworth, Continuous Partial Pressure

Distillation of Essential Oil-bearing Materials: Part I. Bench-scale Tests, College

of Engineering Technical Report, Texas A&I University (1984).

3. W. L. McCabe, J.C. Smith and P. Harriot, Unit Operations of Chemical Engineering,

4th ed., McGraw-Hill, New York (1985).

4. J.H. Perry and C.H. Chilton, eds., Chemical Engineers’ Handbook, 5th ed., McGraw-

Hill, New York (1973).