- Top of page
- Materials and Methods
- Results and Discussion
Because of concerns about Vibrio vulnificus, the U.S. Food and Drug Administration is considering requirements for postharvest processing (PHP) of oysters harvested from the Gulf of Mexico during warm-weather months and intended for raw consumption. As described in the paper, feasible PHP methods for warm-weather-harvested oysters include cool pasteurization, high hydrostatic pressure, and low-dose gamma-irradiation. We estimate that the costs of applying PHP are approximately 5 to 6 cents per half-shell oyster intended for raw consumption. However, most oyster processors have insufficient volumes to cost-effectively install PHP equipment. To assist these smaller operations, central PHP facilities operated by a 3rd party would be needed. A geographic information system analysis that minimized volume-weighted travel distances from each Gulf oyster operation identified 6 optimal PHP facility locations in the Gulf region. Even with the establishment of central PHP facilities, some oyster operations will become unprofitable and be at risk for closure.
- Top of page
- Materials and Methods
- Results and Discussion
In October 2009, the Food and Drug Administration (FDA) announced its intent to require postharvest processing (PHP) of Gulf of Mexico oysters harvested during the warm-weather months that are intended for raw consumption (Kraemer 2009). As of 2013, the state of California is the only state in the United States that requires PHP of oysters harvested from the Gulf of Mexico in summer months. Consumption of raw oysters from the Gulf of Mexico is associated with Vibrio vulnificus illnesses in consumers (U.S. FDA 2009). Vibrio vulnificus is a naturally occurring bacterium found in warm seawater and can be transmitted to humans through the consumption of raw shellfish harvested from waters containing the organism (CDC 2009). It does not normally affect healthy individuals, but persons who are immunocompromised, especially those with chronic liver disease, are at greater risk for being affected by Vibrio vulnificus from oyster consumption (CDC 2009). Although the annual number of reported Vibrio vulnificus illnesses associated with oyster consumption is low, generally in the range of 30 to 35 cases per year (CDC 2011), the incidence rate of death among those individuals who contract the disease is high, approximately 50%.
PHP methods can be applied to raw oysters and essentially eliminate the risk of illness due to Vibrio vulnificus. The focus of this paper is on estimating the costs and assessing the feasibility and economic consequences of requiring PHP if FDA imposes the proposed requirements. The Food Safety Modernization Act, signed into law on January 4, 2011, requires that prior to proposing requirements for PHP, FDA must prepare a report that includes, among other requirements, an assessment of the costs of complying with PHP requirements; analysis of the impact of PHP on sales, costs, and availability of oysters; and an assessment of how PHP requirements may be implemented in the fastest, and safest, and economical manner. The analysis described in this paper helps in addressing some of these requirements. In addition, FDA would need to assess the public health benefits, consult with various state agencies, and compare the requirements with those of other regulated domestic and imported foods.
The methods that have been determined to reduce Vibrio vulnificus to nondetectable levels (<30 MPN/g) are cool pasteurization, high hydrostatic pressure (HHP) processing, low-dose gamma-irradiation, and cryogenic individual quick freezing (IQF) with extended frozen storage. Cool pasteurization is a mild thermal treatment of oysters in the shell, followed by a rapid cooling (Muth and others 2000, 2002a). This process raises the temperature of the oyster enough to kill Vibrio vulnificus bacteria but does not sterilize or cook the oyster. To treat oysters, the oysters are first washed, graded, and then individually banded with rubber bands and loaded onto trays. The trays are loaded onto carts, then hoisted into a tank containing warm water for several minutes, and finally hoisted into a cool water tank for several more minutes. Currently, one operation in south Louisiana uses the cool pasteurization process for raw oysters.
HHP is a method of inactivating microorganisms in foods by subjecting them to very high pressure. This process can be used for both half-shell and shucked oysters. Prior to processing, oysters intended for the raw half-shell market are individually banded using a shrink-wrap band. Workers load banded oysters for both raw half-shell and shucked uses into baskets, which are then conveyed to the ultra-high-pressure processor. The water-filled pressure chamber is then sealed and pressurized for 3 to 5 min using an electric 60-horsepower pump (Avure Technologies 2008). For oysters intended for shucking, the pressure helps release the adductor muscle from the shell, making it easy to remove the oyster from the shell. Currently, 3 operations near the Gulf (2 in Louisiana and 1 in Texas) use the HHP process for raw oysters.
Irradiation of oysters has been approved by FDA as a postharvest process (FDA 2005) and validated by researchers at the Univ. of Florida, although the process is not yet used commercially for oysters. Irradiation involves exposing oysters to ionizing energy, either gamma rays, machine-generated electrons, or X-rays. Gamma rays are more commonly used, specifically from cobalt 60. The gamma rays interact with water and other molecules in the oyster, thereby inactivating bacteria. As of 2013, one irradiation facility operates in the Gulf area, in Florida, but it has not irradiated oysters for the commercial market.
Finally, cryogenic IQF with extended frozen storage is a postharvest process that uses liquid carbon dioxide or nitrogen to freeze oysters on the half-shell. Once treated, the oysters are stored in a freezer for a period sufficient to achieve nondetectable levels of Vibrio vulnificus. One of the benefits of the IQF process is that oysters can be stored from the winter harvest, which yields higher-quality oysters (particularly from the Gulf of Mexico), and then offered for sale during other times of the year. However, we excluded IQF from the analysis in this paper because using IQF for summer-harvested oysters generally results in an unacceptable product from the consumer's perspective. Because oysters spawn in the summer, they are thinner, and the freezing process results in poor color and texture of oysters.
The Gulf States—Alabama, the west coast of Florida, Louisiana, Mississippi, and Texas—account for 60% of oyster harvests in the United States. According to the Natl. Marine Fisheries Service (NMFS), 20.6 million meat-weight pounds of oysters were harvested in 2008 (NOAA Marine Fisheries 2011). Harvest volumes are largest October through March when oyster quality is higher; oyster quality declines when oysters begin to spawn during warmer months. Based on information obtained from state agencies and industry participants, an estimated 40% of Florida-West Coast, 70% of Louisiana, and 75% of Texas oysters harvested from the Gulf in the summer months (April through October) are used for half-shell consumption. Overall, for half-shell and shucked oysters, an estimated 30% of Florida-West Coast, 75% of Louisiana, and 50% of Texas oysters harvested from the Gulf of Mexico in the summer and intended for half-shell consumption are shipped interstate and thus are specifically subject to PHP requirements (Muth and others 2011). Essentially no oysters harvested from Alabama and Mississippi during the summer are used for half-shell consumption.
Applying PHP to all summer-harvested Gulf oysters intended for raw half-shell consumption would require substantially greater PHP capacity than is currently available in the industry (Muth and others 2011, 2012). Although existing oyster processors could increase their PHP capacities by operating more hours per week, installation of additional equipment will likely be needed to meet FDA requirements for PHP. However, smaller oyster processors may lack the resources or volume to install PHP equipment; thus, alternative methods of obtaining PHP services will likely be needed.
The objective of this study was to analyze the costs and economic feasibility of requiring PHP of Gulf State oysters harvested in the summer (April through October) and intended for raw half-shell consumption. We determined the resource requirements for installing and operating PHP equipment, estimated the costs of PHP on a per-oyster basis, and analyzed potentially economically feasible methods of applying PHP to all oysters harvested in the Gulf of Mexico from April through October and intended for raw consumption. The results of the analysis can help guide regulators in determining the most economically feasible and efficient method of implementing PHP with the intent of eliminating illnesses associated with Vibrio vulnificus. The analysis updates and extends data from Muth and others (2000, 2002a) by providing new cost estimates for PHP processes, adding irradiation to the estimation of costs, and considering the heterogeneous nature of oyster processing establishments.
The analysis of the effects of PHP requirements was subject to limitations resulting from several major events affecting the oyster industry. In particular, the Deepwater Horizon oil spill in April 2010 and historic flooding of the Mississippi River in the spring of 2011 resulted in numerous harvest area closures and significant number of death of oysters from freshwater diversions. In addition, the presence of red tide along the Texas coast forced the closure of all oyster harvest areas for the entire state of Texas for the fall 2011 season. Also, the Gulf States each imposed time–temperature requirements in May 2010, which for some states are as restrictive as a 1-h limit from harvest to refrigeration in the summer months.1 All of these events have caused and will cause substantial reductions in oyster harvests for several years into the future. However, use of a PHP technology would allow processors to sell oysters that do not meet the May 2010 time–temperature harvest requirements for raw half-shell consumption.
Ultimately the success of implementing PHP requirements for all summer-harvested Gulf oysters intended for raw half-shell consumption will depend on consumer acceptance of postharvest processed oysters. Safety, sensory, and other quality characteristics will affect the demand for postharvest processed oysters relative to traditional oysters. Multiple sensory studies suggest that consumers of raw oysters and individuals on trained sensory panels could not detect differences between traditional and postharvest processed oysters, with the exception of previously frozen oysters (Balthrop 2001; Andrews and others 2002; Andrews 2003; Andrews and Coggins 2004; Coggins 2004; Otwell and others 2010). However, consumer surveys suggest that most consumers are not willing to buy postharvest processed oysters, and those who are willing to pay about the same amount as they would for traditional oysters (Balthrop 2001; Hanson and others 2003; Posadas and Posadas 2004; Morgan and others 2009). For purposes of this analysis, we assumed the average consumer does not differentiate between traditional and postharvest processed oysters. Further analysis would be needed to determine the full consumer response to postharvest processed oysters if a requirement was imposed.
Materials and Methods
- Top of page
- Materials and Methods
- Results and Discussion
Oyster processors seeking to maximize profits while adhering to requirements for PHP may consider the following options: install and operate PHP equipment within the establishment, obtain PHP services from a private or public operation (if available), or close during the summer months or permanently. Profits from processing oysters can be expressed as
where is profit from processing oysters in year t; , , and are the prices of half-shell, shucked, and shellstock oysters; , , and are the quantities of half-shell, shucked, and shellstock oysters; Wt is the oyster processing wage rate for oyster processing activities (for example, unloading, washing, sorting, shucking, and shipping); Lt is the quantity of oyster processing labor hours; is the price of oyster processing materials (for example, water, ice, and packaging materials); Mt is the quantity of oyster processing materials; is the cost of applying a postharvest process to half-shell oysters (including transportation costs if using a toll processing service); and is the cost of (or reduced cost of) applying a postharvest process to shucked oysters. and include amortized capital equipment and installation costs and annual operating costs based on product volume for each PHP process. In determining whether to close in response to PHP requirements, oyster processors would compare their current profits against the difference between the value of the firm from closing and the discounted present value of profits they would earn from remaining in the industry (Muth and others 2002b). In other words, if they could sell the firm for more than the sum of all future earnings of the firm, then it may make economic sense for them to sell the firm and invest the proceeds in another line of business. Specifically, they would be at risk for closure in year t if
where VLt is the value of the firm when exiting the market (its value if sold) and is the discounted present value of the firm when remaining in the market at the end of the period.
Oyster processors are maximizing profits based on a fixed short-run supply of the primary input, shellstock oysters, and allocating the primary input to 2 types of output: raw half-shell and shucked oysters. Depending on the location and season, shellstock oysters may or may not be available, and the quality of those supplies can vary considerably, thus affecting whether they are used for raw half-shell or shucked. In addition, oyster processors are maximizing profits over the course of the year rather than month to month. Thus, some oyster processors are willing to accept prices that are below their costs for shucked oysters in the summer because they are seeking to satisfy their customers’ needs over the course of the year to retain those customers for the months of the year when shucking yields are higher and, thus, shucking oysters is profitable.
Determining the effects of PHP requirements is complicated by the fact that some oyster processors may opt to sell oysters that have not undergone a PHP process only within the state of harvest. Louisiana and Texas have passed legislation that allows intrastate transport, sale, and consumption of raw oysters that have not been postharvest processed. With this allowance, the raw half-shell market becomes a differentiated product market, but with one product substantially restricted by geographic location compared with the other. Furthermore, establishments that install PHP equipment would likely use the process for both half-shell and shucked oysters to reap the benefits associated with increased yields and reduced shucking costs for shucked oysters, while establishments that would have to rely on toll processing would likely use the process only for half-shell oysters. Thus, there will be differential industry responses because of the possibility of only intrastate shipments and the treatment of half-shell compared with shucked oysters by different industry segments.
The methods used for the analysis of PHP requirements focused on these areas: (1) determining resource requirements and costs of installing and operating PHP facilities in the Gulf of Mexico and (2) analyzing the economic feasibility of PHP of all Gulf oysters harvested in the summer and intended for raw half-shell consumption (such as installing own equipment, using existing or potential private toll-processing facilities, and using potential public toll-processing facilities [or central PHP facilities]). In considering the possibility of central PHP facilities, we conducted a geographic information system (GIS) analysis to determine the general locations that would minimize travel time and costs for operations that currently have no or insufficient treatment capacity.
To estimate the costs of constructing or expanding a plant and operating PHP equipment and to conduct the cost analysis, we conducted in-depth on-site and telephone interviews and obtained detailed information from 3 HHP processors, one cool pasteurization processor, one irradiation processor, and one manufacturer of HHP equipment.2 In addition, we obtained information from local realtors, economic development agencies, and a refrigerated truck transportation company in the Gulf region to estimate costs such as land, construction, insurance, and transportation.
We obtained data on individual oyster operations from the Interstate Certified Shellfish Shippers List (ICSSL), which includes the lists of certified dealers provided by the states to FDA (U.S. FDA 2011). Processing plants that ship oysters across state lines must be certified as interstate shippers. We assumed that the following types of operations from the ICSSL would be required to either install PHP equipment or identify another location that would offer toll-processing services:
- Shellstock shipper: grows, harvests, buys, or repacks and sells shellstock. Shellstock shippers are not authorized to shuck shellfish or to repack shucked shellfish, but they may ship shucked shellfish.
- Repacker: repacks shucked shellfish from a certified shucker-packer into other containers. Repackers may also repack and ship shellstock but may not shuck shellfish.
- Shucker-packer: shucks and packs shellfish. Shucker-packers may act as shellstock shippers or reshippers or may repack shellfish originating from other certified dealers (U.S. FDA 2011).
One additional type of oyster shipper, reshippers, is not likely to install PHP equipment or use toll-processing services because they are not engaged in processing. Instead, we assumed that reshippers would rely on shellstock shippers and shucker-packers to process oysters as required.
We estimated approximate oyster-processing volumes for Gulf oyster-processing establishments by first eliminating processors from the list obtained from the ISSCL that do not handle oysters or only shuck oysters using information obtained by the ISSC from the Gulf state agencies (Moore 2011). We also augmented the ISCCL data with financial information from Dun & Bradstreet (D&B) (www.dnb.com) by matching the establishment name and address with records in the D&B dataset. We then converted the revenue estimates into estimated annual numbers of oysters processed by each establishment using average price estimates.3 To account for the fact that many shippers handle products other than oysters; thus, their revenue estimates represent other types of products. We scaled back the volumes to account for other products using percentages estimated by calibrating the estimated volumes for operations on the shippers list to 2008 harvest volumes as reported by NMFS.4 We also adjusted the volumes produced by existing PHP processors by subtracting their postharvest processed volumes from their total volumes to obtain an estimate of the remaining volume of oysters that would need to undergo PHP.
Resource and cost estimation
Each of the PHP methods is associated with increased capital equipment, labor, or energy requirements and potential revenue changes due to changes in the type or nature of the product sold. We calculated the resources and costs associated with installation of cool pasteurization and HHP methods, but relied on toll-processing cost information from the irradiation facility in Florida to calculate the costs of using irradiation as a PHP method. For the cool pasteurization and HHP methods, we used the information collected during the industry interviews to develop typical estimates of capital equipment costs (and life of capital equipment) and costs of labor, energy, materials, and insurance for representative size operations. These assumptions included purchasing land (estimated at $70000 for a small facility or $175000 for a large facility) and building additional plant space (estimated at $150 per square foot or $1614 per square meter, with an expected life of 20 y). Capital equipment and other initial costs were annualized assuming a 7% interest rate and added to annual operating costs to develop a total annual cost estimate for each postharvest process.
To estimate the costs of insurance for additional plant space and PHP equipment in existing establishments and for the entire facility and equipment in central PHP facilities, we obtained estimates of the annual cost per million dollars of assets from Internet searches, economic development organizations, and oyster processors. We used an estimated value of 3% of the value of the insured assets. For example, $1000000 in insured assets would result in a $30000 annual premium.
For calculations requiring conversion of oyster volumes, we applied several assumptions obtained through discussions with industry participants (Muth and others 2011). Specifically, we assumed 250 oysters per 100-pound sack at harvest; 7 pounds of oyster meat per 100-pound sack of oysters average over the course of the year, which equates to approximately 36 oysters per meat-weight pound; 4 pounds of oyster meat per 100-pound sack of oysters in the summer, which equates to approximately 62 oysters per meat-weight pound; and 60% of Gulf-harvested oysters are sold for half-shell use and 40% are sold for shucking over the course of a year. After applying these assumptions, we calculated total and per-oyster costs of PHP for shucked and half-shell oysters using the data provided by oyster processors and PHP vendors, as described below.
Cost estimates for the cool pasteurization process are based on the following volumes: 18000 sacks of oysters per year for a small process and 145600 sacks per year using holding tanks with capacities of 7500 gallons (hot tank) and 5500 gallons (cold tank) for a large process (Fahey 2010). Capital equipment requirements for cool pasteurization include a boiler, chilling and condensing unit, computer-monitored hot and cold exchange unit, holding tanks (7500 gallons for the hot-water tank and 5500 gallons for the cold-water tank), conveyers, hoists for lifting oysters in and out of water tanks, an ultraviolet water purification system, stainless steel racks, and delivery and installation including plumbing and electrical hookups. Estimates of the costs of plant expansion to house the equipment were calculated assuming 200 square feet (18.58 m2) for the small process and 1750 square feet (162 m2) for the large process. For both the small and large processes, capital equipment and installation costs were estimated by applying a net inflation factor of 1.31 obtained from the Bureau of Labor Statistics for the period 1999 to 2009 (2010 was not yet available) to original cost estimates provided by AmeriPure in 1999 (Muth and others 2000).
Capital equipment costs (including installation) and plant expansion costs were amortized assuming a 20-y life and 7% interest rate. Current estimates for operating costs—water, electricity, natural gas, labor, replacement parts, and maintenance—were added to banding costs and adjusted for shucking labor savings to develop total annual operating costs. In addition, a licensing fee of $0.0125 per oyster was included.5
High hydrostatic pressure
The sole equipment manufacturer for HHP equipment, Avure, produces 4 sizes of machinery that can process oysters: 100-liter (L) horizontal machine operating at 11 cycles per hour with 120 shell-weight pounds per cycle (requiring floor space of 12 by 12 feet, or 13.3 m2), 320-L vertical machine operating at 12 cycles per hour with 450 shell-weight pounds per cycle (requiring floor space of 30 by 20 feet, or 55.7 m2), 350-L horizontal machine operating 12 cycles per hour with 500 shell-weight pounds per cycle (requiring floor space of 50 by 20 feet, or 92.9 m2), and 687-L horizontal machine operating 10 cycles per hour and with 700 shell-weight pounds per cycle (requiring floor space of 40 by 30 feet, or 111.5 m2). Capital equipment requirements for HHP are the HHP unit and enclosure, chiller, compressor, overhead rail system, conveyers, hoists, and delivery and installation costs, including electrical hookups.
Licensing fees for HHP are built into the capital equipment costs and, thus, are not separately incurred on a per-oyster basis. Plant expansion costs were estimated assuming the minimum required square footage would be twice the footprint of the HHP equipment. However, the 320-L vertical system requires 23 feet, or 7 m, of vertical clearance, which would be difficult in many facilities, in contrast to the horizontal system, which is 6 to 7 feet (2 m) in height. Thus, plant expansion costs may be higher than estimated for installing a vertical process.
Avure provided estimates of the base equipment costs; additional costs for installation, rail system, conveyors, and building expansion; and operating costs per shell-weight pound, including labor, electricity, water, building expansion, conveyors, and depreciation costs (using a straight-line method). To provide consistency in estimating the costs of plant expansion per square foot and amortizing costs using a 7% interest rate, we decomposed the per-pound operation costs provided by Avure and then reconstructed the plant expansion, installation, and annual per-oyster costs of HHP. We estimated plant expansion costs by multiplying $150 per square foot times twice the square footage requirements provided by Avure. Capital equipment costs were included as provided by Avure. We estimated additional equipment and installation costs assuming that the costs are 10% of capital equipment costs based on detailed information provided by HHP processors that use a 215-L machine and a 350-L machine each. Per-oyster operating costs were calculated by subtracting our estimate of the portion of Avure's per-oyster operating costs attributable to plant expansion, capital equipment, and installation and adding back our annualized estimate of each of these portions of costs assuming a 20-y life for plant expansion, 10-y life for capital equipment and installation, and a 7% interest rate.6 We then adjusted the per-oyster operating costs to account for banding costs for half-shell oysters and shucking labor savings and increased yields for shucked oysters.7
Following these calculations, we compared the resulting cost estimates to cost estimates calculated using detailed information provided by HHP processors based on their recent experience installing HHP processes. The estimates based on the data from the HHP processors were somewhat higher than, but generally similar to, the estimates provided by Avure. The cause of the differences is unknown but could be due to a variety of factors, including differences in the wages and energy prices, imprecision in the method we used to deconstruct Avure's cost estimates, or differences in assumptions used.
Large quantities of oysters can be irradiated quickly within packaged boxes. Oyster processors would ship packaged oysters to an irradiation facility for treatment and incur costs for banding oysters prior to packaging, toll-processing charges, and transportation to and from the irradiation facility. Irradiation would be the last step in the process before oysters are introduced into commerce. In trials conducted by the irradiation facility, oysters were cleaned, packaged, labeled at the oyster processing facility and then shipped to the irradiation facility on pallets in refrigerated trucks.8 The irradiation facility does not need to hold oysters, because they can process an entire truckload in only 1 h. Thus, the oysters are transported to and from the irradiation facility on the same truck.
Central PHP facilities
Processors that use central PHP facilities would incur toll-processing costs and transportation costs to and from their closest central PHP facilities. We estimated an approximate toll-processing cost by first annualizing total initial investment costs including land (1 acre or 4046 m2 for a small facility and 2.5 acres or 10117 m2 for a large facility), site development costs ($99750 for a small facility and $266000 for a large facility), construction costs (7500 square feet or 696 m2 for a small facility and 20000 square feet or 1858 m2 for a large facility at $150 per square foot), process validation costs of $15000, and PHP equipment purchase and installation costs. We then calculated the annualized total initial investment cost on a per-oyster basis and added it to the per-oyster operating cost estimated for privately owned facilities.
To estimate the costs of transporting oysters to obtain PHP services, we assumed that each processor using the service would make 84 round trips during the summer months to obtain PHP services (3 trips per week for 28 wk). Annual costs are the number of miles per year (84 trips times each round trip to a central PHP facility) multiplied by the estimated per-mile cost for refrigerated truck transportation of $2.43 per mile obtained from oyster processors and a refrigerated truck transportation company in the Gulf region.9 We then calculated transportation costs per oyster by dividing the costs per year by the estimated volume of half-shell oysters shipped interstate from April through October for each facility. We assumed that oyster processors would only apply PHP to half-shell oysters intended for interstate shipment based on the expectation that all states would allow intrastate shipments of untreated oysters.
Economic feasibility of PHP
To conduct the economic feasibility analysis, we first compared the estimated production volume of each oyster processing facility with the processing capacities of each size of PHP equipment to determine if the facility has sufficient production volumes to enable installation of the equipment. Because most processors have insufficient production volumes, we then conducted a GIS analysis to determine the optimal locations for central PHP facilities. We then evaluated whether all oyster processors that would need to install PHP equipment or obtain PHP services would continue to operate profitably.
In conducting the GIS analysis, we assumed that oysters would be shipped from a processor location to a central PHP facility to allow for preprocessing activities (cleaning, sorting, and banding) at the processor location. Oysters would then be either shipped back to the processor location for final packaging and order fulfillment or directly to a buyer. Oyster processors would, therefore, incur costs for refrigerated shipping to and from the central PHP facility in addition to the costs of PHP services. The analysis was based on the assumption that all summer-harvested Gulf half-shell oysters shipped interstate would be treated using cool pasteurization or HHP. We assumed that a central PHP facility would have at most a monthly treatment capacity equivalent to the highest capacity HHP processor operating 4800 h per year or 2 of the highest-capacity cool-pasteurization units also operating for 4800 h per year.
To determine the optimal locations for central PHP facilities, we used ESRI's Network Analyst extension within ArcMap with the following optimization criteria: minimize the travel distance from the original establishment to the central PHP facility using major highways and require that the central PHP facility be within a 4-h drive from the original establishment to allow time for loading and unloading and for drivers to return the same day. The specific Network Analyst analysis used for this was Location-Allocation. This analysis required 3 data inputs: the locations of the oyster establishments, the locations of the PHP facilities, and the road network along which transport would occur. The road network used was ESRI's streets, the standard road network included with ESRI's data bundle in version 10.0. Both the oyster establishments and the PHP facilities were “snapped” to the closest network road, thereby allowing the distance along the network between the 2 to be calculated. The network possessed measures of both distance (miles) and time (minutes), allowing the analysis to specify a maximum proximity threshold (in this case 4 h). We ran scenarios with different numbers of sites that minimized the travel time for all establishments assigned to a PHP facility. After the scenarios were run, we calculated the total volume of oysters per site and then selected the optimal number of sites that maximized processing volume while not exceeding the processing capacity.