Anda di halaman 1dari 8

Journal of Food Engineering 83 (2007) 414421 www.elsevier.

com/locate/jfoodeng

A quality cost model for food processing plants


Aurora Zugarramurdi a,*, Maria A. Parin a, Liliana Gadaleta a, Hector M. Lupin b
b a INTI Mar del Plata/CIC, Southern Regional Centre, Marcelo T. de Alvear 1168, 7600 Mar del Plata, Argentina Fishery Industries Division, FAO/DANIDA Training Programme, Viale delle Terme di Caracalla, Roma I-00100, Italy

Received 13 October 2006; received in revised form 12 March 2007; accepted 16 March 2007 Available online 28 March 2007

Abstract A HACCP-based system is a recognized food safety management program aiming at the control of all the factors aecting food safety. It is also possible to add factors related to food quality. To evaluate the eectiveness of a quality system, a realistic estimate of quality costs is essential. The purpose of this work is to develop a mathematical model for the calculation of the costs associated with a specic quality level due to HACCP-based system implementation. Experimental results obtained at Argentinean hake freezing plants (Merluccius hubbsi) are presented and compared with those calculated with the proposed model. The proportion of variance explained by the model was 0.903 for total quality costs; proving its optimum performance. 2007 Elsevier Ltd. All rights reserved.
Keywords: HACCP; Quality costs; Food; Fish freezing plant

1. Introduction The HACCP system constitutes a food safety management system based on a scientic, systematic, rational, multi-disciplinary, and cost-eective approach, controlling safety problems in food processing, especially for high-risk foods such as meat, poultry or shery products (Deodhar, 2003; Huss, 1995; Jensen & Unnevehr, 2000; Ziggers, 2000; Zugarramurdi, Parin, Gadaleta, Carrizo, & Lupin, 1999). Properly applied, there is no other system or method able to provide the same degree of safety and quality assurance at a lower cost (Huss, 1994; Jensen & Unnevehr, 2000). Few studies have addressed the costs and benets of HACCP implementation in the food business. As a consequence, it is dicult to evaluate the extent to which costs and benets to businesses act as an incentive/disincentive to the further adoption of HACCP within the food sector (Henson, Holt, & Northen, 1999). Many food processing plant managers feel that quality programs decrease plant productivity, thereby making such

Corresponding author. Tel./fax: + 54 223 489 1324. E-mail address: auroraz@inti.gov.ar (A. Zugarramurdi).

programs a costly luxury. This may be true for an initial implementation period, but it is not indenitely true (Bonnell, 1994). A study of HACCP adoption in the UK dairy industry shows that rms are adopting HACCP in order to meet customer as well as legal requirements and to gain improvements in operating eciency (Henson et al., 1999). To evaluate the eectiveness of a HACCP-based system, a realistic estimate of quality costs is essential. Quality costing activities can be carried out to gain top management commitment, direct improving eorts, and, above all, estimate the benets of quality improvement (Huss, 1995; Saita, 1991). Quality costing is the rst step for preparing a case for Total Quality Management (TQM) initiative. Moreover, a realistic estimation of quality costs is an essential element of any TQM initiative. However, only a minority of organizations uses formal quality costing methods because quality costs are hard to measure. This work aims to develop a mathematical model for the calculation of the costs associated with a specic quality level due to HACCP-based system implementation. Experimental results obtained at Argentinean hake freezing plants (Merluccius hubbsi) are presented and compared with those calculated with the proposed model.

0260-8774/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jfoodeng.2007.03.029

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421

415

2. Model and assumptions Quality-related costs are divided into: prevention costs, appraisal costs and failure costs (PAF model), so stated by Feigenbaum (1974). The proposed Quality Cost Model consists of two submodels. The controllable costs sub-model is made up of prevention costs (Pi) and appraisal costs (Ai). Controllable costs are those that the manager or decision maker can control directly so as to assure product quality. They are associated with investments in the non-conformances prevention and product appraisal for conformances to requirements. Each component of the controllable costs is aected by coecient X that depends on quality levels. The sub-model of failure costs consists of those expenses incurred by the rm as a consequence of all its activities associated with defective products failing to meet quality requirements, prior or after delivery to the customer. In this category both internal and external failures are included. Failure costs were evaluated as direct company losses. Quality costs are analysed following the Pareto principle: few factors account for the largest portion of costs (Sandholm, 1987). Generally speaking, in order to estimate quality costs, good judgment is necessary to avoid excessive attention to minor items that, even if severely over-or underestimated, will not have a signicant eect on the overall estimate. The model assumes that quality costs can be calculated from eleven dierent components, as shown in Table 1. To calculate the quality costs associated with a specic product quality level, the relationship between the raw material and nal product quality should be known. The chemical, physical, biological and sensory quality characteristics of the raw material should be studied together with the product quality. The proposed model was developed considering the full quality range within the limits required by sanitary and safety regulations for both raw materials and products.
Table 1 List of categories and elements of quality costs Controllable costs Prevention costs P1. Design, development and implementation of a quality assurance plan P2. Quality training programs for suppliers and production personnel P3. Hygiene and sanitation of the plant P4. Preventive maintenance and additional supervision Appraisal costs A1. Receipt and control of incoming material A2. Sampling and laboratory analysis A3. In-process inspection External failure costs F4. Claims, rejected and recalled products Resulting costs Internal failure costs F1. Scraps, reprocessing or spoilage F2. Low labour productivity and low process yield F3. Inecient use of plant capacity

It is also important to recognize the variations in raw material quality levels and their eects on process yield and labour productivity in the food processing plant under analysis. Quality, market, and production parameters used in the proposed model are dened in Table 2. 2.1. Controllable costs sub-model 2.1.1. Prevention costs In food processing plants, the most representative prevention costs are the following: P1. Design, development and implementation of a quality assurance plan. Quality planning costs, C P 1 , include the expenses arising from designing the quality plan and audit system, from the design and development of quality measurements as well as from the purchase of control equipment. Properly applied and established, the great advantage of the HACCP system is that the main eort of quality assurance is directed towards the Critical Control Points and away from endless nal product testing (Huss, 1994; Jensen & Unnevehr, 2000). C P 1 can be estimated as follows: C P 1 XP 1 I F =dK 1 Before HACCP is implemented, food plants must follow prerequisite programs such as GMPs, and a Sanitary plan (US Food & Drug Administration & Institute of Food Technologists, 2001). The coecient XP 1 can be estimated as a percentage of the xed investment (IF), and it represents about 35% of the xed investment for a food plant working at very high
Table 2 Quality, market and production parameters Quality parameters Qm = raw material quality (dimensionless parameter) Qp = product quality (dimensionless parameter) Y Qm = yield for optimum quality level Qm (kg product/kg raw material) X Qm = productivity for optimum quality level Qm (kg product/h-worker) Nscp = number of sanitation control points Nccp = number of critical control points P Qm = purchase price of raw material (US$/kg) P Qp = selling price for quality level Qp (US$/kg) P Qp = selling price for optimum quality level Qp (US$/ kg) d = number of working days per year K = daily production (kg product/day) Ri = raw material sampling inspection rate (kg raw material/h) S = average labour rate for trained workers (US$/h) IF = total xed investment (US$) X Qm = productivity for quality level Qm (kg product/ h-worker) Y Qm = yield for quality level Qm (kg product/kg raw material) L = total labour cost (US$/kg product)

Market parameters

Production parameters

416

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421

quality levels (Gall, 2000; Golan, Ralston, Frenzen, & Vogel, 2000; Jensen & Unnevehr, 2000; US Food & Drug Administration & Institute of Food Technologists, 2001). P2. Quality training programs for suppliers and production personnel. Quality training costs, C P 2 , include the costs of developing, implementing, operating and maintaining formal quality training programs for both suppliers and production personnel. Thereby, size and quality uniformity is achieved for each lot. This cost will also increase savings in sampling in the reception step for well-trained suppliers. Expression (2) was used to calculate C P 2 . The XP 2 coefcient could be dened as a percentage of total labour cost. With technically appropriate training programs, safe food production should be accompanied by protability. For food processing plants, the maximum training cost usually spent is around 4050% of the total labour cost (Food Standards Agency, 2002; Jensen & Unnevehr, 2000). C P 2 XP 2 LQp 2

activities and total hygiene and sanitation cost. In the food industry, overall labour comprises the bulk of costs, i.e., over 90% (Gould & Gould, 1993; US Food & Drug Administration & Institute of Food Technologists, 2001), with detergents, disinfectants and cleaning equipment taking up only a minor part. P4. Preventive maintenance and additional supervision. In order to evaluate preventive maintenance and additional supervision costs, C P 4 , it is necessary to consider the costs of adjustments and repairs that assure the optimum operation of the equipment. Preventive maintenance is commonly regarded as a necessary element of high productivity, and even more for product safety. It avoids costly product recalls or product defects. Besides, an extra supervision is required for prevention. To estimate C P 4 , expression (4) is used: C P 4 XP 4 I F =Kd XP 5 LQp 4 Preventive maintenance costs can be estimated as a percentage of the additional investment required for implementing a HACCP-based system. Using economic engineering techniques, maintenance costs are estimated as a percentage of the xed capital investment (Jelen & Black, 1983; Zugarramurdi et al., 1995). In the food sector, maintenance costs range from 2% to 6% of IF (Zugarramurdi et al., 1995; Parin & Zugarramurdi, 1994). Taking into account the additional investment (35% of xed investment), introduced with expression (1), coecient XP 4 resulted as 0.10.3% of IF. Additional supervision cost is estimated by taking a xed percentage of direct labour costs based on company experience. Using economic engineering techniques, supervision in food plants is usually estimated at 1020% of direct labour costs (Jelen & Black, 1983; Parin & Zugarramurdi, 1994; Zugarramurdi et al., 1995). This cost is introduced in expression (4) by the XP 5 coecient. Coecient XP 5 is estimated as a variable percentage of labour costs. It can be estimated as 510% of direct labour costs in food processing plants (Colatore & Caswell, 2000, Golan et al., 2000). 2.1.2. Appraisal costs These comprise inspection and testing costs and ensure that the products, parts, and raw materials conform to quality requirements. A1. Receipt and control of incoming material. These are the costs related to the in-plant (internal) inspection and testing of the main raw material and to the salaries and payroll charges of inspection personnel. The control of incoming material is the rst critical control point for a food plants quality assurance system. Once the critical control points have been determined, a sampling system including sample location, size, frequency, selection methods, testing analysis, and reporting is established in accordance with inspection norms and standards specied for each food product (Gould & Gould, 1993; Hubbard, 1996).

P3. Hygiene and sanitation of the plant. Cleaning is one of the most important activities in the prevention of failure costs in food processing plants (ECOLAB, 1997; Gould & Gould, 1993; Huss, 1994; Jensen & Unnevehr, 2000; US Food & Drug Administration & Institute of Food Technologists, 2001; Zugarramurdi, Parin, & Lupin, 1995). Cleaning costs, C P 3 , include the inputs (detergent, disinfectants) and labour costs incurred to clean, as determined by the established procedures. It is a function of the total labour cost (Dunsmore, Tomlinson, & Ashley, 1983). Factory hygiene as well as personal hygiene and sanitation are usually dened as sanitation control points in the prevention of product contamination with microorganisms, lth or any other foreign material during processing. The standards of hygiene required to avoid such problems vary. In plants where packaged processed products are produced, safety requirements are stringent, whereas those for fresh chilled food with a short shelf life and cooked before consumption are less demanding (Huss, 1994). The corresponding values of hygiene and sanitation costs were calculated with expression (3). In literature, there is a wide variation in estimated costs of implementing HACCP, evincing the inherent uncertainties and wide range of possible assumptions, such as the number of sanitation control points. Therefore, an additional 1 percent was assigned to account for the expenses associated with each control point of the SSOPs plan (Jensen & Unnevehr, 2000). C P 3 XP 3 L1 0:01 NscpQp =z 3

The coecient XP 3 was introduced in expression (3) to contemplate hygiene labour costs. In the food industry, studies have shown that hygiene and sanitation costs can be up to 20% of total labour costs (Colatore & Caswell, 2000; Golan et al., 2000). Coecient z (decimal) accounts for the relationship between the labour included in hygiene and the sanitation

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421

417

The cost of inspection and testing of the main raw material also includes inspection at the suppliers premises by the purchasers sta. The cost of inspection, C A1 , can be computed with expression (5) C A1 XA1 S =Y Qm Ri Qp 5

Eq. (5) shows that this cost is inversely proportional to the raw material sampling inspection rate (Ri) and to the overall yield. The analysis of the parameter XA1 indicates the inspection activities performed in the actual plan in relation to those that should be done for the ideal quality. A2. Sampling and laboratory analysis. Sampling and laboratory cost includes the costs related to the tests needed to evaluate the quality of purchased raw material, semi-nished or nished products as well as the depreciation cost of specialized quality control and assessment equipment. This item also includes all the in-plant laboratory costs (sampling and testing) and any outside testing services used. As temperature is the single most important parameter in HACCP in the food industry, the costs related to measuring, recording and temperature controlling (equipment, labour cost and maintenance) are also included (Henson et al., 1999; Jensen & Unnevehr, 2000; Zugarramurdi et al., 1995). The sampling and laboratory cost is determined as a percentage of labour cost as shown in Eq. (6). This percentage can range from a low percentage to as much as 15%, depending upon plant complexity, implementation time and product quality; all this introduced in expression (6) by the XA2 coecient (Golan et al., 2000; Jensen & Unnevehr, 2000; Ollinger, Moore, & Chandran, 2004). C A2 XA2 LQp 6

and lack of an adequate HACCP-based quality plan, capable of preventing losses in yield, and product reprocessing, that reduced corporate productivity and protability (Bonnell, 1994; Valdimarson, 1992). They are estimated as the decrease in selling price (less quality), labour ineciency (less productivity) and raw material (less yield). In the long run, a superior quality level allows the producer to step up to a higher value market position. F1. Scraps, reprocessing or spoilage. Scraps, reprocessing or spoilage cost (C F 1 ) includes losses resulting from the dierence between an optimum selling price and a reduced price obtained by the producer due to non-conformance for quality reasons. This cost is a direct loss and should include all the material, processing costs and labour costs incurred during production. In particular, it includes the quality aspects connected to yield, wastage (raw materials, intermediate products, nal products, energy, labour, facilities), substandard products (not reprocessed) and markdowns (also called downgrading) due to poor quality. An example of this type of cost is the so-called overall shrink. Shrink is dened as the loss of inventory due to spoilage, deterioration, water loss or uncontrollable acts. In the food industry, fresh raw materials average overall shrink is about 10% (Somsen & Capelle, 2002; Zugarramurdi et al., 1995). In the sh industry, the overall shrink for frozen sh accounts for 1% (Zugarramurdi et al., 1995). This cost is evaluated by expression (8), adapted from Taguchi function (Saita, 1991). The dierence between the best selling price (optimum quality) and the current price is used as a parameter. C F 1 P Qp P Qp 1 Qp
2

A3. In-process inspection. The in-process inspection cost (C A3 ) includes the personnel engaged in the in-process evaluation of product conformance to quality requirements. C A3 increases as extra supervision is required to reach high levels of product quality and can doubled the labour costs in food processing plants (Jensen & Unnevehr, 2000; Zugarramurdi et al., 1995). It is a function of the number of critical control points (Nccp), since these are the stages requiring monitoring, record keeping and procedures for out of control situations. C A3 XA3 NccpLQp 2.2. Failure quality cost sub-model When a quality assurance system is not correctly applied, internal and external defects arise, incurring in lost opportunity costs. 2.2.1. Internal failure costs These are the costs arising within an organization due to non-conformance or defects at any stage of the quality loop. Internal failures result from low raw material quality 7

F2. Low labour productivity and low process yield. These costs include the cost of personnel and idle facilities resulting from product defects, disrupted production schedule, and yield losses due to poor raw material quality. These costs can be estimated as follows:
C F 2 S 1=X Qm 1=X Qm P Qm 1=Y Qm 1=Y Qm

F3. Inecient usage of plant capacity. This cost should be accounted for in factories using poor raw material quality, as it leads to a reduction in production capacity because of the decrease in the overall yield of the whole process. This cost can be estimated as follows:
C F 3 P Qp Y Qm Y Qm =Y Qm

10

2.2.2. External failure cost This is dened as the cost arising after delivery to a customer due to non-conformity or defects. This type of failure is probably the most costly of all. Considering the food processing industry, this could apply to: claims, rejected and/or returned products, etc.

418

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421

This cost can be calculated from the number of claims made by the clients. This number depends on the product price, since, if high, the consumer is prepared to formalize the claim while if low, the claim is not recorded (Sandholm, 1987). A model with an exponential function of the eorts accomplished in the quality evaluation is proposed, in accordance with the function put forth by Saita (1991); therefore, C F 4 is calculated as an exponential function of appraisal costs (CA): C F 4  expC A C A C A1 C A2 C A3 11

where is based on the number of claims and on the product price. To estimate value, the rm must know the characteristics of the demand function associated with the product. Reductions in external complaints are relevant for diminishing production costs, improving prot margin and increasing the market share of the rm. 2.3. Total quality costs The summation of all prevention, appraisal and failure costs represents the variation of total quality costs (TQC) per unit of product, as a function of product quality level (Qp) X X X TQCQp C P i Q p C Ai Qp C F i Qp 12 In principle, the increase of expenditure in prevention and appraisal costs should decrease failure costs (external and internal), and there should be a point at which total quality costs is at its lowest value. This study focuses on the calculation of the direct benets from the implementation of a HACCP-based system. There are other side eects that generally lead to higher benets. Also, TQM techniques can further increase the indirect benets, but the scope of the model hereby presented is on the lower limits of the benets to be gained by enforcing a HACCP for food safety assurance. 3. Case study. Hake freezing plants The model hereby presented was applied to manual sh processing plants in Mar del Plata city, Argentina. These plants elaborate interleaved blocks of frozen hake llets for export and already have proper GMP and SSOPs. Usually these plants operate 270 working days per year. Technical and economical data were elaborated from previous works and from experience in actual sh freezing plants (Crupkin et al., 1996; Giannini, Parin, Gadaleta, Carrizo, & Zugarramurdi, 2001; Montaner, Gadaleta, Parin, & Zugarramurdi, 1995; Zugarramurdi & Parin, 1988; Zugarramurdi, Parin, Gadaleta, Carrizo, & Lupin, 2004; Zugarramurdi, Parin, Gadaleta, & Lupin, 2000).

To calculate the costs associated with a specic quality level, a relationship between raw material and nal product quality was obtained. Physical, chemical, biological, and sensory quality characteristics of fresh whole hake and frozen interleaved hake llets were jointly analysed in a dimensionless basis. In most cases analysed, the value of the parameter corresponding to the best quality obtained (maximum value) was used as the denominator in order to obtain each dimensionless parameter. When the related variables had a negative correlation slope, the minimum value of the variable was subtracted from both the numerator and denominator. Applying this method, the scale ranged from 0 for the worst quality to 1 for the best quality. Dierent intervals are dened in Table 3 as a function of raw material quality. Least squares regression analysis has been applied to the collected data, and the following relationship between raw material and product quality for iced and frozen white sh was found (Zugarramurdi et al., 2004): Qp 0:832Qm 0:033 R2 0:928; Students t test; P < 0:01 3.1. Controllable costs Each component of the controllable costs is aected by coecients X that depend on quality level. The values of these coecients are shown in Table 4 and explained below. 3.1.1. Prevention costs Fish processing plants already complying with GMPSSOPs only need as much as one percent of IF for planning and implementing the quality system so as to reach a very good quality (NOOA, 1993). Three percent was used for the XP 1 value corresponding to optimum quality. In this study, plant capacities felt within the 2050 raw material
Table 3 Raw material quality levels Quality level Poor Fairly good Good Very good Extra quality Qm 100 624 2548 4971 7294 P95

13

Table 4 Coecients values for sh freezing plants as a function of quality level Quality level Extra quality Very good Good Fairly good Poor XP 1 0.03 0.01 0.005 0 0 XP 2 0.5 0.3 0.1 0 0 XP 3 0.20 0.15 0.10 0.05 0.05 XP 4 0.004 0.003 0.002 0.001 0 XP 5 0.05 0.04 0.02 0.01 0 XA1 1 0.5 0.3 0.1 0 XA2 0.12 0.09 0.06 0.03 0 XA3 0.02 0.01 0.005 0 0

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421

419

tonne per day range. A sh freezing plant of 20 t/day capacity has a IF of approximately US$2.5 million (Montaner et al., 1995). The cost of training and XP 2 coecient descend from 0.5 to 0.1 when the level of product quality is decreased from extra quality to good (Jensen & Unnevehr, 2000). In the shery industry, overall labour comprises the bulk of costs (average 84.2%), with detergents and disinfectants accounting for only a minor part (average 10.6%) and cleaning equipment for even less (average 5.2%) (Dunsmore et al., 1983); thus coecient z was xed at 0.842 in Eq. (3). Hygiene labour costs in sh freezing plants were found to be up to 20% of the total labour costs. In general, the SSOPs plans have seven sanitation control points (SENASA, 1996). Coecient XP 4 was determined between 3% and 10% of maintenance costs. Maintenance cost for sh freezing plants is estimated at 4% of IF. The maximum X for additional supervision cost was a 5% of the direct labour costs for optimum quality, and consequently coecient XP 5 was estimated as shown in Table 4 (Zugarramurdi et al., 1995). 3.1.2. Appraisal costs In order to apply the proposed model to frozen sh processing plants, the sampling plan used was assumed to be designed according to regulations proposed by the Canadian Quality Management Program (Fisheries & Oceans Canada, 1989) for sh processing plants at standard quality level. The sample size is determined according to the number of sh in the lot that is related to the weight of the sh. As above mentioned, the parameter XA1 indicates the inspection activities performed in the actual plan in connection with the activities that should be done to reach optimum quality, in which case X can be as high as 1. In Argentina, a typical requirement for inspection activity is about 75 kg raw material per hour (R). In the sh processing sector average labour rate is around 4 US$/h. Sampling and laboratory costs in sh processing plants can rise up to 12% of labour cost for maximum product quality (Zugarramurdi et al., 1995). The number of process critical control points (Nccp) in frozen sh processing plants is usually 1 (reception of raw material) (National Marine Fisheries Service, 1990; SENASA, 1996). Labelling and packing are not taken into account as critical control points when fraud is not considered as a HACCP factor. Another critical control point usually contemplated in freezing processing plants is food ingredients, which does not apply to this product. It should also be noted that the end-use of this frozen product is cooking. 3.2. Failure costs Process yield and labour productivity for frozen hake llets have been analysed in previous works (Zugarramurdi and Parin, 1988; Montaner, Parin, Zugarramurdi, Lupin, 1994a, 1994b; Crupkin et al., 1996; Zugarramurdi et al.,

1995; Gadaleta et al., 2003; Zugarramurdi et al., 2004) as a function of raw material or product quality. The linear relationship between lleting yield (Y) and product quality for hake processing found by Zugarramurdi et al. (2004) was used; and is shown in Eq. (14) Y Qm 17:27Qp 32:39 R2 0:978; Students t test; P < 0:01 14 Expression (15) shows the linear relationship between labour productivity or workers training level (X) and product quality found by Zugarramurdi et al. (2004). X Qm 25:76Qp 26:78 R2 0:925; Students t test; P < 0:01 15 For interleaved frozen blocks of hake llets, yield and productivity increase 7% and 12%, respectively, when the quality level of the product is increased from good to very good. Raw material price ranged from 0.25 to 0.45 US$/kg while selling price ranged from 1.4 to 1.85 US$/kg as a function of quality. Then, expressions (8)(10) were used to estimate internal failure costs due to losses or decreases in sale price (less quality) and ineciencies in labour and raw material. 3.3. Total quality costs Regression analysis has been applied to the controllable costs (CC), failure costs (FC), and total quality costs (TQC) per unit of product for dierent levels of quality resulting from the application of the proposed model. Regression equations for the model were obtained tting polynomial curves to the model values by least squares estimations and were plotted in Fig. 1. For frozen hake processing plants analysed, it can be observed that when product quality level increases from fairly poor to very good, controllable costs increase from
1.2

0.8 Quality Costs US$/kg

0.6

0.4

0.2

0 0 0.2 0.4 0.6 Product Quality 0.8 1

Controllable Costs

Failure Costs

Total Quality Costs

Fig. 1. Total quality costs, failure costs and controllable costs as a function of product quality for frozen hake processing plants.

420

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421 Colatore, C., & Caswell, J. A. (2000). The costs of HACCP. Implementation in the seafood industry: A case study of breaded sh. In L. J. Unnevehr (Ed.), The economics of HACCP: Costs and benets (pp. 4568). St. Paul: Eagan Press. Crupkin, M., Montecchia, C., Montaner, M. I., Parin, M. A., Gadaleta, L. B., Beas, V., et al. (1996). Inuence of the reproductive cycle of Argentinian hake (Merluccius hubbsi) of llet yield and operating costs. Journal of Aquatic Food Product Technology, 5(3), 2939. Deodhar, S. Y. (2003). Motivation for and cost of HACCP in Indian food processing industry. Working Paper, Indian Institute of Management Ahmedabad, Available from www.ijeb.com/journals/downloads/ dec03/mfchifpi.pdf. Dunsmore, D. G., Tomlinson, P., & Ashley, R. J. (1983). Cleaning practices in the Australian seafood processing Industry. Food Technology in Australia, 35(12), 566570. ECOLAB (1997). Sanitation process evaluation and control system. Brochure. Ecolab Inc., Available from www.ecolab.com/Initiatives/ foodplantoperations/Images/SanitationSPECS Cat Sht R1.pdf. Feigenbaum, A. V. (1974). Total quality control (1st ed.). New York: McGraw-Hill. Fisheries and Oceans Canada (1989). QMP specications. Sampling plan and acceptance numbers for the inspection of sh/shellsh. Quality Management Program. Submission Guide. Canada. Food Standards Agency (2002). Meat (Hazard Analysis and Critical Control Points) (England) Regulations 2002. Regulatory Impact Assessment, Available from www.food.gov.uk/multimedia/pdfs/ regulation_haccp.pdf. Gadaleta, L., Parin, M. A., Lupin, H. M., & Zugarramurdi, A. (2003). Calidad de la materia prima y su inuencia sobre la calidad y costos de n en productos pesqueros. Revista de Tecnolog a e Higiene de produccio los alimentos: Alimentaria, 343, 4147. Gall, K. (2000). National Seafood Industry HACCP Implementation n Program, Available Survey Report. New York Sea Grant Extensio from www.seagrant.sunysb.edu/SeafoodTechnology/seafoodpubs.htm. Giannini, D., Parin, M. A., Gadaleta, L., Carrizo, G., & Zugarramurdi, A. (2001). Inuence of raw material quality on quality of iced and frozen white sh products. Journal of Food Quality, 24, 527538. Golan, E. H., Ralston, K. L., Frenzen, P. D., & Vogel, S. J. (2000). The costs, benets and distributional consequences of improvements in food safety: The case of HACCP. In L. J. Unnevehr (Ed.), The economics of HACCP: Costs and benets (pp. 149169). St. Paul: Eagan Press. Gould, W., & Gould, R. (1993). Total quality assurance (2nd ed.). Baltimore: CIT Publications. Henson, S., Holt, G., & Northen, J. (1999). Costs and benets of implementing HACCP in the UK dairy processing sector. Food Control, 10, 99106. Hubbard, M. R. (1996). Statistical quality control for the food industry (2nd ed.). New York: Chapman and Hall. Huss, H. H. (1994). Assurance of seafood quality. FAO Fisheries Technical Paper No. 334. Huss, H. H. (1995). Quality and quality changes in fresh sh. FAO Fisheries Technical Paper No. 348. Jelen, F. C., & Black, J. (1983). Cost and optimization engineering (2nd ed.). New York: McGraw-Hill Book Company.. Jensen, H. H., & Unnevehr, L. J. (2000). HACCP in pork processing: Costs and benets. In L. J. Unnevehr (Ed.), The economics of HACCP: Costs and benets (pp. 2944). St. Paul: Eagan Press. Montaner, M. I., Gadaleta, L., Parin, M. A., & Zugarramurdi, A. (1995). Estimation of investment costs un sh processing plants. International Journal of Production Economics, 40, 153161. Montaner, M. I., Parin, M. A., Zugarramurdi, A., & Lupin, H. M. (1994a). Requerimientos de insumos de la industria pesquera: I. a e Higiene de los alimentos: Materia prima. Revista de Tecnolog Alimentaria, 253, 1924. Montaner, M. I., Parin, M. A., Zugarramurdi, A., & Lupin, H. M. (1994b). Requerimientos de insumos de la industria pesquera: II. Mano

5% to 73% of TQC and failure costs decrease from 95% to 27% of TQC. It can also be observed that when product quality level improves from good to very good, TQC decreases 35%. At the same time, there is a zone with lower values around very good level of product quality. In order to validate the model, controllable costs, failure costs and total quality costs in actual frozen hake processing plants were also collected. The points in Fig. 1 correspond to experimental values. To evaluate how this model tted the experimental values, the proportion of variance accounted for in the dependent variable by the model was computed. The proportion of the variance explained was 0.865 (for controllable costs), 0.949 (for failure costs) and 0.903 (for total quality costs). This proves the high performance of the model described. 4. Conclusions To evaluate the eectiveness of the HACCP-based system implemented in food processing plants, a mathematical quality cost model is proposed. The model consists of two controllable costs and resulting cost sub-models. The model assumes that quality costs can be calculated from eleven different components. Each component of controllable costs is aected by coecients that depend on the quality level and on quality, market, and production parameters. Failure costs were estimated as direct company losses. The model is applied to sh processing plants and compared with experimental values obtained from real plants. A very good regression coecient for total quality costs was obtained, showing that the model explains the 90.3% of the experimental values from actual sh freezing processing plants. Acknowledgements The authors express their thanks for the nancial support by INTI, CONICET, CIC, UNMDP, and FAO. Much appreciated was the assistance of the shing industry associations of Mar del Plata (Argentina), which allowed the authors to make the necessary industrial assays. AZ is aliated with National Institute of Industrial Technology (INTI) and is a Scientic Research Member of the Scientic Research Commission of the Buenos Aires Province (CIC). MAP and LG are Members of the Research and Development Sta of the National Council of Scientic and Technical Research (CONICET). HML is aliated with Fishery Industries Division, FAO Fisheries Department. AZ, MAP, and LG are Members of the Sta of the College of Engineering, National University of Mar del Plata, Argentina. References
Bonnell, A. D. (1994). Quality assurance in seafood processing: A practical guide (1st ed.). New York: Chapman and Hall.

A. Zugarramurdi et al. / Journal of Food Engineering 83 (2007) 414421 a e Higiene de los alimentos: Alimentaria, de obra. Revista de Tecnolog 254, 8185. National Marine Fisheries Service (1990). HACCP Regulatory Model Raw Fish. Model Seafood Surveillance Project. Oce of Trade and Industry Services. Pascagoula, MS. NOOA (1993). The report of the Model Seafood Surveillance Project Draft. National Oceanic and Atmospheric Administration, US Department of Commerce, National Marine Fisheries Services, Oce of Trade and Industry Services. Washington, DC. Ollinger, M., Moore, D., Chandran, R. (2004). Meat and poultry plants food safety investments: Survey ndings. Economic Research Service, USDA, Technical Bulletin No. 1911. Electronic report, Available from www.ers.usda.gov/publications/tb1911/tb1911_researchbrief.pdf. Parin, M. A., & Zugarramurdi, A. (1994). Investment and production costs analysis in food processing plants. International Journal of Production Economics, 34, 8389. . Strategia e costi, 1st ed. ISEDI Saita, M. (1991). Economia della qualita Petrini Editore, Torino, Italia, pp. 915. Sandholm, L. (1987). Reducing quality costs to improve export earnings. International Trade Forum, 2631. n del Sistema SENASA (1996). Manual de procedimientos. Aplicacio ca lisis de Riesgos y Puntos crticos de control. Artegra HACCP. Ana Impresos, Buenos Aires, Argentina. Somsen, D., & Capelle, A. (2002). Introduction to production yield analysis A new tool for improvement of raw material yield. Trends in Food Science and Technology, 13, 136145. US Food and Drug Administration and Institute of Food Technologists (2001). Analysis and evaluation of preventive control measures for the

421

control and reduction/elimination of microbial hazards on fresh and fresh-cut produce, Available from www.cfsan.fda.gov/~comm/ift3toc.html. Valdimarson, G. (1992). Developments in sh processing Technological aspects of quality. In H. H. Huss et al. (Eds.), Quality assurance in the sh industry (pp. 169183). Elsevier Science Publications. Ziggers, G. W. (2000). HACCP, vertical coordination and competitiveness in the food industry. In L. J. Unnevehr (Ed.), The economics of HACCP: Costs and benets (pp. 269284). St. Paul: Eagan Press. Zugarramurdi, A., Parin, M. A., & Lupin, H. M. (1995). Economic Engineering applied to the shery industry. FAO Fisheries Technical Paper No. 351. Zugarramurdi, A., Parin, M. A., Gadaleta, L., Carrizo, G., & Lupin, H. M. (1999). The economics of HACCP. Producers point of view: Increasing benets and reducing costs. Presented at the Advanced Seminar on Implementation and Economics of HACCP in the Fishery Industry, Hirtshal, Denmark. Zugarramurdi, A., & Parin, M. A. (1988). Economic comparison of manual and mechanical hake lleting. Engineering Costs and Production Economics, 13, 8995. Zugarramurdi, A., Parin, M. A., Gadaleta, L., Carrizo, G., & Lupin, H. M. (2004). The eect of improving raw material quality on product quality and operating costs: A comparative study for lean and fatty sh. Food Control, 15, 503509. Zugarramurdi, A., Parin, M. A., Gadaleta, L., & Lupin, H. M. (2000). The economics of HACCP application in Argentine sh products. In L. J. Unnevehr (Ed.), The economic of HACCP: Costs and benets (pp. 403412). St.Paul: Eagan Press.