Aquaculture 259 (2006) 315 327 www.elsevier.

com/locate/aqua-online

A bioeconomic evaluation of a commercial scale recirculating finfish growout system An Australian perspective
Paul N. De Ionno a,, Graeme L. Wines b , Paul L. Jones a , Robert O. Collins a
b a School of Life and Environmental Sciences, Deakin University, P.O. Box 423, Warrnambool, Victoria 3280, Australia School of Accounting, Economics and Finance, Deakin University, P.O. Box 423, Warrnambool, Victoria 3280, Australia

Received 14 December 2005; received in revised form 17 May 2006; accepted 24 May 2006

Abstract This study, based on 3 years of commercial data, presents the results of an economic analysis of a 20-tonne per annum (TPA) commercial recirculating aquaculture system (RAS) facility located in Warrnambool, Victoria, Australia. Based on the assumptions of the analysis, results highlight the non-viability of the facility, with a 10-year projected negative cumulative cash flow of $648,038, and negative net present value (NPV) of $707,546. Economies of scale were assessed by the development of economic models for hypothetical 50-TPA and 100-TPA facilities, based on the actual figures obtained from the 20-TPA case study. These analyses highlighted marginal viability for the 50-TPA facility (with a ten-year projected cumulative cash flow of $1,030,300; negative NPV of $167,651 and internal rate of return (IRR) of 11.75%), and an economically viable 100-TPA facility (with a ten-year projected cumulative cash flow of $3,176,750; NPV of $522,200 and IRR of 21.03%). Sensitivity analysis highlighted that the greatest gains to be realised in improving profitability were those associated with increasing the productive capacity of the facility, increasing the sale price of the product, and decreasing the capital costs of RAS facilities. Contradictions between the results from the present study to similar studies clearly highlight a need for further economic analyses of commercial RAS facilities, using commercial data sets and standard economic analysis procedures. 2006 Elsevier B.V. All rights reserved.
Keywords: Recirculating aquaculture systems; Economics; Murray cod; Production; Profitability

1. Introduction The aquaculture industry has benefited from over four decades of research aimed at developing technically viable production systems (Kazmierczak and Caffey, 1995). Improved nutrition, species selection, disease prevention, water quality management and systems development have allowed not only widespread establishment

Corresponding author. Tel.: +61 3 5563 3061; fax: +61 3 5563 3462. E-mail address: pauld@deakin.edu.au (P.N. De Ionno). 0044-8486/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2006.05.047

of pond systems, but also the emergence of recirculating aquaculture systems (RAS) (Kazmierczak and Caffey, 1995). RAS are a relatively new technology designed for holding and growing a wide variety of aquatic species and are defined as production units that recycle water by passing it through filters that remove metabolic and other waste products (Kazmierczak and Caffey, 1995). In comparison to traditional aquaculture practices (i.e. pond and cage culture), RAS offer more independence from the external environment (i.e. increased levels of control), which can provide a basis for improved risk management (Rawlinson, 2002).

316

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

The commercialisation of RAS technology has only begun to show signs of maturity in recent times and the industry is widely accepted as being in its infancy in comparison to other aquaculture production techniques. For example, in Australia in 200102, reported aquaculture production in RAS amounted to 549 tonnes, with an approximate value of AUD $6.5 million, this representing approximately 1.24% of the total Australian aquaculture market share by yield and 0.88% by value (Love and Langenkamp, 2003). As a result of the infancy of the RAS industry, there has been a great deal of emphasis on biological and engineering developments, but a shortage of useful research which combines this information with the economics of RAS. Hence, these developments have not generated widespread economic success for commercial producers (Kazmierczak and Caffey, 1995; Honda, 1998). Previous studies have demonstrated the following areas to have had the greatest potential economic impact on the profitability of RAS: biological variables (feed conversion ratio, FCR; survival and growth rates); operating cost variables (cost of feed; labour; power and oxygen); engineering performance variables (filtration efficiency; level of intensification); system and fixed cost variables (capital costs; economies of scale; total production); and revenue (sale price). However, due to the difficulties and costs of conducting economic experiments on commercial scale operations (Kazmierczak and Caffey, 1995), the majority of these previous studies have been based on bioeconomic models incorporating hypothetical data sources. In many cases, models have been developed assuming best practice husbandry techniques, maximum production and sale of all output once stock have completed their growout period. As most experienced aquaculturists understand, these kinds of best-case scenarios are rare in practice. Accordingly, most information summarised from these studies should only be taken as a general and theoretical guide. Powless (1998) criticised the validity of bioeconomic modelling and theoretical data sources; stating that, Paper fish, those that are grown in models, don't distinguish between mediocre and superior feeds. They don't require good water quality and they never die. Unfortunately, there is no market for paper fish. These statements highlight the need for more economic analyses based on real commercial data sets for RAS facilities. This will provide the industry and potential investors with a more accurate understanding of the economic viability of RAS ventures, hence potentially reducing the number of future economic failures. Thus, the primary objective of this study is to analyse RAS from a bioeconomic framework and provide in-

dustry with a detailed economic analysis of an actual, commercial scale RAS facility. It is hoped that these results will provide a quantitative technique to assist managers to understand and interpret many of the interrelated processes in RAS and the economic constraints faced by profit seeking producers. 2. Materials and methods 2.1. Study site and data collection All data for this study was collected from a commercial RAS facility located in Warrnambool, Victoria,
Table 1 Summary of all parameters monitored between July 1, 2002 and June 30, 2005 (Years 1 to 3) at the WTF facility Income Net fish sales ($, excluding wholesaler's commission and market dues) Number of fish sold Average weight of fish sold (g) Total production (kg) Costs Capital and infrastructure costs RAS facility Shed Vehicle Operational costs Depreciation Electricity Equipment Feed Freight/packaging Gas/diesel General administration, office expenses Stock insurance Legal fees Maintenance Marketing/advertising Oxygen Permits Phone/internet Plumbing/pipes/fittings Salaries Chemical/cleaning products Security Stock purchases Travel/conferences Vehicle costs Water quality, fish health assessment Biological variables Feed conversion ratio (FCR) Feed consumption Mortality Weight gain

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

317

Australia. This facility, known as the Warrnambool Trout Farm (WTF) facility, is jointly owned and operated by Deakin University and Warrnambool Trout Farms Pty Ltd, as part of the Sustainable Aquaculture Systems program. Data was collected for this study between July 1, 2002 and June 30, 2005. This period incorporates the construction and start up phase (July 1, 2002 to May, 2003) and the commissioning period (defined as the period to obtain maximum specified standing stock and/ or feed rate, i.e. June, 2003 to January, 2005), through to maximum output (January, 2005 onwards). Table 1 summarises the wide range of data that was collected throughout this period in order to obtain an accurate and valid bioeconomic evaluation of the facility. All data is presented annually (i.e. Years 1, 2 and 3) to accurately assess and compare financial performance on an annual basis. All financial data presented in the study is recorded in Australian dollars (AUD). 2.2. System specifications A basic schematic of the WTF facility is summarised in Fig. 1, whilst Table 2 summarises its general specifications. The facility has cultured both Murray cod (Maccullochella peelii peelii Mitchell), and Short finned eels (Anguilla australis Richardson) throughout the first three years, but its main focus, output and future plans

Table 2 Summary of the specifications of the WTF RAS facility Supplier Australian Aquaculture Systems Pty. Ltd. Location Warrnambool Trout Farm Date commissioned June 2003 Species cultured Murray cod, Maccullochella peelii peelii Short finned eel, Anguilla australis Maximum specified feed rate 100 kg/day Maximum specified output 20 tonnes per annum (TPA) Standing stock at maximum 12.5 tonnes/55,000 fish output Stocking density in culture 100150 kg/m3 tanks

have centred on Murray cod, with eels phased out of the facility during Year 2. 2.3. Income and expenditure Market price (i.e. income) was closely monitored throughout Years 1 to 3. Data was plotted graphically using linear regression analysis so as to identify any possible trends between market price per unit and unit size. Using these data, a figure for market price per unit ($/kg) was derived in order to project long term income requirements and for projections of larger scale output projections.

Fig. 1. A basic schematic of the WTF commercial RAS facility.

318

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

A detailed summary of all expenditure (i.e. capital and operating costs) was maintained throughout Years 1 to 3 of the project. Major cost categories were highlighted and presented graphically in order to identify key variables that affected profitability of the WTF facility. 2.4. Economic analysis techniques An annual summary of the income and expenditure of the WTF facility was summarised and presented in cash flow statements and profit and loss statements. However, for the purpose of this study only the results from the cash flow statements will be presented as this is widely recognised as the preferred technique for analysing long term, high risk investments of this type. Using the collected data, projections for the facility were carried out to Year 10, as it is unlikely that an aquaculture enterprise would be an attractive investment opportunity if it were not profitable after ten years (Staniford, 1989; Head and Watanabe, 1995). From these financial statements, the following measures of financial performance were derived: 2.4.1. Cumulative cash flows and NPV A measure that represents the total, gross amount of the net cash flows (i.e. inflows less outflows) over a specific period. Cash inflows are netted off against cash outflows over the period, and hence a positive value indicates that inflows exceed outflows. While cumulative cash flows provide an indication of a positive cash position, net present value (NPV) extends the analysis by taking into account the timing of the cash flows and the time value of money. NPVanalysis applies the time value of money to cash inflows and outflows over the investment project's life so that management can evaluate a project's benefits and costs at one point in time (Larson et al., 2002). It explicitly recognizes that a dollar received today (present value) is worth more than a dollar to be received at some future time (O'Rourke, 1996). NPV is computed by discounting the future net cash inflows at the project's required rate of return, and then subtracting the initial amount invested (Larson et al., 2002). Hence, a positive NPV highlights that the present value of the net cash inflows to be received over the project's life exceeds the amount of the initial investment, and hence is an indicator of economic viability. 2.4.2. Internal rate of return (IRR) Internal rate of return is a rate used to evaluate an investment's feasibility which reflects the rate of return the project earns (Petty et al., 1996; Larson et al., 2002). Mathematically, IRR is the discount rate that yields an

NPV of zero for an investment (Larson et al., 2002). Hence, a project evaluated according to IRR is accepted if its IRR is greater than or equal to the required rate of return (Petty et al., 1996). A minimum IRR or discount rate of 15% was used in the present study and has often been used as a criterion for high risk investments of this type, with a higher IRR required as investment risk, market uncertainty and the cost of capital increases (O'Rourke, 1991; Head and Watanabe, 1995; Van Wyk, 2000). The 15% discount rate adopted in the present study can be generally defined as incorporating a 5% risk free rate (i.e. the return from investing in a bank or term deposit), plus a risk premium of 10%. Nevertheless, it can be highlighted that the exact discount rate chosen for an NPVanalysis is not specifically an issue when the IRR is also calculated, as the IRR for a project indicates the rate at which a project is considered to become acceptable in an economic sense. For example, if the calculated IRR for a project is 10%, then this indicates acceptability, at least in an economic sense, for investors who would expect a return of at least 10% for that individual project. It also enables comparison with the return available on a risk-fee investment. 2.4.3. Payback period Payback period can be defined as the expected time period required to recover the initial investment amount (Larson et al., 2002). Companies desire a short payback period to increase return and reduce risk (Larson et al., 2002). For example, a payback period less than a project's useful life or a pre-specified period indicates investment feasibility (Head and Watanabe, 1995). 2.4.4. Discounted payback period The discounted payback period is calculated with the application of discounted future net cash inflows (i.e. 15% in the present study), while the payback period is calculated using non-discounted future net cash inflows. 2.5. Assumptions of the economic analysis As with all economic analyses, various assumptions must be made in order to simplify the study and maintain the focus of the analysis to the appropriate areas. The assumptions used in this study were developed using a conservative approach and are summarised in Table 3. All income, operating costs and biological data were based on actual data obtained during the first three years of operation of the WTF facility. Goods and Services Tax (GST) was excluded from all costs as these costs are refundable for this kind of operation in accordance with Australian taxation laws.

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

319

Table 3 Summary of assumptions used in the development of ten-year financial projections for the WTF 20-TPA RAS facility and hypothetical 50-TPA and 100-TPA facilities Variable Capital and infrastructure costs Depreciation rates (i.e. useful life) 20-TPA Actual data 15 years for 50% of plant and equipment 30 years for 50% of plant and equipment 5 years for motor vehicles Actual data $0 Excluded Years 410 100% $12.50/kg (whole fish, 1 kg+) Years 410 1.2 (Years 410) $1600/tonne (Years 410) Actual data Actual data Excluded Pre-tax Owner investment no borrowings 15% 50-TPA $1,100,000 15 years for 50% of plant and equipment 30 years for 50% of plant and equipment 5 years for motor vehicles 12 months (Year 1) $0 Excluded Year 2 25% Year 3 75% Years 410 100% $12.50/kg (whole fish, 1 kg+) 1.2 $1400/tonne $100,000 p.a. (inc. on costs) $60,000 p.a. Excluded Pre-tax Owner investment no borrowings 15% 100-TPA $1,625,000 15 years for 50% of plant and equipment 30 years for 50% of plant and equipment 5 years for motor vehicles 12 months (Year 1) $0 Excluded Year 2 25% Year 3 75% Years 410 100% $12.50/kg (whole fish, 1 kg+) 1.2 $1300/tonne $175,000 p.a.(inc. on costs) $100,000 p.a. Excluded Pre-tax Owner investment no borrowings 15%

Construction phase Salvage value Land value Output

Sale price FCR Feed costs Labour costs Electricity costs GST Tax rate Capital source Discount rate

Depreciation rates were determined in accordance with standard industry practice. A salvage value of $0 was used for all plant and equipment at the expiration of the depreciation period for conservative reasons, although it could be expected that a facility of this size would obtain some salvage value at the expiry of the project. Land value was excluded from the analysis in order to focus on the profitability of the RAS operation itself. Land value can be identified as an investment in its own right, hence potentially exhibiting reasonable capital growth over the ten-year period, especially due to land improvements created by the construction of a commercial RAS facility. Nevertheless, any capital appreciation in land over the period of the RAS operation does not represent an economic return from the actual RAS facility. All income was recorded on a pre-tax basis to focus the results on the performance of the facility. This assumption eliminated the differing tax rates that are available depending on the structure of the operation (e.g. partnership, company, trust, etc.) and differing tax laws between countries and states. In particular, tax concessions for primary producers in Australia considerably reduce the amount of tax potentially payable. Finally, the assumption of owner investment was applied (i.e. no borrowings, interest, etc.). This again focuses the analysis on the performance of the RAS facility and minimises the effects of incorporating highly variable external factors.

2.6. Economies of scale To assess and compare the effects of economies of scale on profitability, economic models were developed for hypothetical 50-TPA and 100-TPA facilities. The economic analysis techniques used were identical to those used for the WTF 20-TPA facility. It is believed that the data obtained for the WTF 20-TPA facility during its first three years of operation will add strength to the hypothetical data and assumptions developed for these larger scale facilities. As with the WTF 20-TPA facility model, various assumptions must be made for the hypothetical 50-TPA and 100-TPA facilities. These assumptions are summarised in Table 3. Income data, based on the results from the analysis of market price for the 20-TPA facility (refer to Section 3.1), were applied to these models in order to determine the long term income projections for both the hypothetical 50-TPA and 100-TPA facilities. Quotations from two commercial RAS suppliers (including the supplier of the WTF 20-TPA facility) offering systems of similar configuration, were used in order to estimate the capital costs for these larger scale facilities, with the average of the two prices used. The majority of operating costs for the 50- and 100TPA facilities were developed based on the costs incurred and experiences at the WTF facility. Noticeably,

320

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

Table 4 Summary of parameters used in sensitivity analyses and the variations applied to them in order to assess their potential impacts on the financial performance of the WTF 20-TPA facility and the hypothetical 50- and 100-TPA facilities Variable Biological variables Feed conversion ratio (FCR) Operating cost variables Feed costs Labour costs Electricity costs System and fixed cost variables Capital costs Total production/output Revenue Sale price Variation 1.51.8 +/ 20% +/ 20% +/ 20% +/ 20% +/ 20%

and 100% from Year 4 onwards. These output rates are based on the data collected and experiences obtained whilst operating the WTF facility, where it took 3 years to develop a standing stock base capable of producing a consistent output of 1-kg size fish. The remainder of the assumptions (i.e. depreciation rate, salvage value, land value, feed conversion ratio, GST, tax rate, capital sourcing and discount rate) were identical to those used for the WTF 20-TPA analysis (Table 3). 2.7. Analysis of key variables affecting RAS profitability In order to compare the effects of different variables on profitability, sensitivity analysis (or what if? analysis) was used in this study to highlight areas where improvements to RAS may have the greatest performance and economic impacts (Losordo and Westerman, 1994). Table 4 summarises the key bioeconomic variables that were identified and subsequently included in the analyses based on the results of the WTF facility (Section 3.1) and from comparisons with the findings from previous studies. Each sensitivity range was set highlighting a realistic best case/worse case scenario for each parameter. The feed conversion ratio (FCR) variable is a measure used to determine the rate at which the stock is converting feed to growth. FCRs at the WTF facility for cod batches grown from fingerling to a market size of 1 kg have averaged 1.2 and hence this value was included into the economic analyses (Table 3). However this FCR is purported to be quite low for commercial RAS in Australia, particularly considering it included

$10/kg vs. $15/kg

significant feed and energy cost discounts were included in the 50- and 100-TPA models as a result of cheaper rates available with increased requirements (based on industry quotations, Table 3). Labour costs were estimated based on the combination of requirements and experiences of operating the WTF facility and liaison with industry (Table 3). Based on the construction phase and output rates during the commissioning period for the WTF 20-TPA facility, a 12-month construction phase (i.e. Year 1) was assumed for the 50- and 100-TPA facilities, whilst output volumes during commissioning through to maximum output were allocated at 25% for Year 2, 75% for Year 3

Table 5 Summary of production from the WTF facility between July 1, 2002 and June 30, 2005 (Years 1 to 3) Year 2 (July 1, 2003 to June 30, 2004) Production Standing stock gain Sales Murray cod (whole fish) Average weight (p.a.) Average weight (Year 2 and Year 3 combined) Short finned eels (live product) Total production Income Murray cod (whole fish) Sale price (average p.a.) Sale price (Year 2 and Year 3 combined) Eels (live product) Total Income 4292 kg 1687.30 kg 950 g 3278.60 kg 9257.90 kg $28,284.98 $16.76/kg $21,310.90 $49,595.88 Year 3 (July 1, 2004 to June 30, 2005) 6263 kg 11,510.50 kg 590 g 620 g 0 17,773.50 kg $118,298.76 $10.28/kg $11.11/kg 0 $118,298.76

No production was achieved in Year 1, due to construction of the facility. Income excludes wholesaler's commission and market dues. Standing stock gain represents production in the form of growth of stock on hand within the period.

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

321

Fig. 2. Relationship between unit weight and market price for Murray cod sold from the WTF facility to wholesale fish markets in Melbourne and Sydney in 2004.

mortality and wasted feed, thus indicating sound management practices and system performance. As a result, FCRs of 1.5 and 1.8 were used in the sensitivity analyses. These values are regularly obtainable and common in commercial RAS operations, and it is unlikely that an average FCR of much less than 1.2 would be a realistic goal for commercial RAS. Key operating cost variables (feed, labour and electricity costs) were identified as the main operating costs of the WTF facility (Section 3.1), and subsequently were included in the sensitivity analyses. All costs were varied by +/ 20% to enable a direct comparison. System and fixed cost variables (capital costs and total production) were varied by +/ 20% as these variations are also commonly evident in commercial RAS. Finally, the large fluctuations in

sale price evident in today's market (Section 3.1) were addressed by comparing the financial performance at both $10/kg and $15/kg. 3. Results 3.1. Economic analysis of WTF 20-TPA RAS facility Table 5 summarises total production from the WTF facility during the first 3 years of operation. These results highlight that, as cod sales increased between Years 2 and 3 (from 1687 kg to 11,510 kg) and as average weight per unit decreased (950 g, Year 2 vs. 590 g, Year 3), the market price for the product (sold as whole fish) decreased by 38.7% ( $16.76/kg Year 2 vs. $10.28/kg Year 3).

Fig. 3. Relationship between unit weight and market price for Murray cod sold from the WTF facility to wholesale fish markets in Melbourne and Sydney in 2005.

322

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

Table 6 Summary of expenditure from the WTF facility during Years 13 Year 1 Capital and infrastructure costs Vehicle WTF shed construction WTF RAS facility Total Capital and Infrastructure Costs Year 2 Year 3

9,818.18 147,430.55 357,820.00 515,068.73

0.00 0.00 0.00 0.00

0.00 0.00 0.00 0.00

Operating costs Electricity RAS 0.00 35,715.68 39,487.37 Electricity WTF 0.00 0.00 3306.83 Nursery Equipment 345.77 7052.65 8157.74 Fish food 1710.50 24,851.45 37,831.20 Freight/packaging 60.00 733.18 5100.77 Gas/diesel 0.00 641.77 1158.92 General administration, 176.14 188.89 145.09 office expenses Insurance stock 0.00 3465.00 11,550.00 Maintenance 69.55 1476.57 2828.03 Marketing/advertising 0.00 0.00 0.00 Oxygen 0.00 1612.91 1216.93 Permits 0.00 0.00 830.91 Phone/internet 75.66 1051.30 1286.55 Plumbing/pipe/fittings 0.00 2073.65 2676.54 Salaries 0.00 100,000.00 100,000.00 Chemicals/cleaning 1298.66 2952.06 6386.27 products Security 51.99 398.63 450.63 Stock purchases 7977.27 66,382.95 24,000.00 Travel/conferences 0.00 0.00 627.27 Vehicle expenses 736.36 3609.12 2450.43 Water quality, fish 0.00 1157.44 0.00 health assessment WTF land 0.00 140.00 0.00 improvements Total operating costs 12,501.92 253,503.26 249,491.47 Total cash outflows $527,570.65 $253,503.26 $249,491.47 Depreciation not included.

relationship between sale price and sale size evident in 2004 (Fig. 2), the long term projection for Murray cod unit sale price was set at $12.50/kg. Table 6 summarises all costs incurred at the WTF facility during Years 1 to 3. As the system was stocked in mid-June 2003, one month of operating costs was incurred in Year 1, with the remainder being infrastructure costs. Annual depreciation costs amounted to $24,769 and were calculated using the straight line method (i.e. equal depreciation costs p.a. over the asset's life), in accordance with the rates stated in the assumptions of the analysis (Table 3). Fig. 4 highlights and identifies the major operating costs incurred during operation of the WTF facility. These results clearly highlight that labour, electricity and feed were the three largest operating cost variables (35%, 15% and 14% of total operating costs respectively), followed by depreciation (9%), and stock (9%). Hence labour, electricity, feed and depreciation (as a function of capital costs) were incorporated into the sensitivity analyses to assess the key variables affecting profitability. Fig. 4 also highlights that the variable other contributed 14% of the total operating costs of the facility. This variable was not included in the sensitivity analyses as it represents a combination of the 17 remaining operating costs that were not separately included in Fig. 4 (Table 6). Analysis of the key financial performance indicators of the WTF facility is summarised in Table 7. These results highlight that the facility will be performing at a large loss for both short term and long term projections, with a projected negative cumulative cash flow of $648,038 at Year 10. Application of the 15% discount rate (i.e. the

Linear regression analysis highlighted a strong decline in market price with decreasing unit size throughout 2004 (R2 = 0.89; p < 0.0001, Fig. 2). However, this relationship appeared not to be evident in 2005 (R2 = 0.04; p = 0.291, Fig. 3), with the market price appearing to have stabilised. These trends highlight an unpredictable market for Murray cod in recent times, hence making it difficult to accurately predict a long term sale price (i.e. income) for the product. Nevertheless, Table 5 also highlights that the average market price received for all product sold from the WTF facility during the first 3 years of operation was $11.11/kg, at an average unit size of 620 g. Based on the long term plans of the WTF facility to sell a minimum of 1-kg units (as whole fish), and based on the positive

Fig. 4. Summary of major operating costs (as % of total costs) for the WTF 20-TPA RAS facility.

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

323

minimum rate of return, or IRR, that potential investors would normally apply to a long term, high risk investment of this type) produced a negative NPV of $707,546 for the 10-year period (Table 7), hence indicating that the investment would be strongly rejected. As a result of the large negative NPV, a negative IRR was incurred and payback and discounted payback periods would never be achieved. 3.2. Economies of scale hypothetical 50-TPA facility Analysis of the financial performance of the 50-TPA facility summarised in Table 7 highlights that this facility could be regarded as marginally viable with a projected cumulative cash flow at Year 10 of $1,030,300. However, the application of the 15% discount rate produced a negative NPV of $167,651 over the 10-year period (Table 7). This negative NPV indicates that the investment is marginally viable, although it would still be rejected by the majority of potential investors. But with a positive IRR of 11.75% and a payback period of 6 years and 8 months (Table 7), the facility could be regarded as approaching economic viability for a high risk investment. As a result of the negative NPV, the discounted payback period for the facility was not achieved within the 10-year forecasting period, hence further emphasising the marginal viability of the 50-TPA facility. 3.3. Economies of scale hypothetical 100-TPA facility Analysis of the financial performance of the 100-TPA facility summarised in Table 7 highlights that the facility could be regarded as a viable investment with a projected cumulative cash flow at Year 10 of $3,176,750 and a positive NPV of $522,200. Further analysis of financial performance highlighted an IRR of 21%, payback period of 5 years and 4 months, and a discounted payback period of 7 years and 7 months (Table 7).
Table 7 Summary of major financial performance indicators for the WTF 20TPA facility and the hypothetical 50- and 100-TPA facilities Profitability measure Cumulative cash flow Net present value Internal rate of return (pre-tax) Payback period Discounted payback period 20-TPA $648, 038 $707, 546 <0 N/A N/A 50-TPA $1,030,300 $167,651 11.75% 6 years, 8 months N/A 100-TPA $3,176,750 $522,200 21.03% 5 years, 4 months 7 years, 7 months

Table 8 Summary of results of sensitivity analyses of measures of financial performance (after 10 years unless otherwise indicated) for the WTF 20-TPA facility Net present value ($) FCR 1.5 1.8 Feed costs 20% +20% Labour costs 20% +20% Electricity costs 20% +20% Capital costs 20% +20% Production 20% +20% Sale price $10/kg $15/kg 737,747.23 767,947.63 683,386.51 731,707.15 644,629.33 770,464.33 684,666.88 730,426.78 604,533.08 810,560.58 864,840.58 550,253.08 864,840.58 550,253.08 IRR (%) <0 <0 <0 <0 <0 <0 Payback period N/A N/A N/A N/A N/A N/A Discounted payback period N/A N/A N/A N/A N/A N/A

<0 <0 <0 <0 <0 <0 <0 <0

N/A N/A N/A N/A N/A N/A N/A N/A

N/A N/A N/A N/A N/A N/A N/A N/A

Discount rate 15%. Results from Production and Sale price sensitivity analyses produced identical results.

3.4. Sensitivity analysis of key variables affecting RAS profitability The results of the sensitivity analyses of key variables affecting the profitability of the WTF 20TPA facility are summarised in Table 8. These results clearly highlight that none of the key variables had any major effect on improving the viability of the facility in terms of producing a positive NPV. The results of the sensitivity analyses of key variables affecting the profitability of the hypothetical 50-TPA facility are summarised in Table 9. In contrast to the WTF 20-TPA facility, these results highlight the marginal viability of the facility, with payback periods and positive IRRs obtained at all sensitivities. However, a positive NPV (and hence a positive discounted payback period) was only obtained when capital costs were reduced by 20% ($52,348, Table 9), production was increased by 20% or sale price was increased to $15/kg ($323,645, Table 9). Table 9 also summarises the results of the sensitivity analyses of potential key variables affecting the profitability of the hypothetical 100-TPA facility. In contrast to the 20- and 50-TPA facilities, these results

324

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

Table 9 Summary of results of sensitivity analyses of measures of financial performance (after 10 years unless otherwise indicated) for the hypothetical 50TPA and 100-TPA facilities 50-TPA NPV ($) FCR 1.5 1.8 Feed costs 20% +20% Labour costs 20% +20% Electricity costs 20% +20% Capital costs 20% +20% Production 20% +20% Sale price $10/kg $15/kg IRR (%) 9.90 7.99 13.18 10.27 13.61 9.84 Payback period 7 Yr 1Mth 7 Yr 6 Mth 6 Yr 5 Mth 6 Yr 11 Mth 6 Yr 4 Mth 7 Yr 1 Mth Discounted payback period N/A N/A N/A N/A N/A N/A 100-TPA NPV ($) IRR (%) 19.12 17.17 22.52 19.51 22.90 19.13 Payback period 5 Yr 7 Mth 5 Yr 10 Mth 5 Yr 1 Mth 5 Yr 6 Mth 5 Yr 1 Mth 5 Yr 7 Mth Discounted payback period 8Yr 3 Mth 9 Yr 0 Mth 7 Yr 2 Mth 8 Yr 1 Mth 7 Yr 1 Mth 8 Yr 3 Mth

258,724.51 349,797.34 94,793.42 240,509.95 72,220.00 263,083.36 110,392.68 224,910.69 52,348.32 387,651.68 658,948.44 323,645.08 658,948.44 323,645.08

353,065.23 183,929.98 657,508.68 386,892.28 689,205.92 355,195.04

12.87 10.61 16.17 11.75 0.40 20.75 0.40 20.75

6 Yr 6 Mth 6 Yr 11 Mth 5 Yr 11 Mth 7 Yr 5 Mth 9 Yr 10 Mth 5 Yr 4 Mth 9 Yr 10 Mth 5 Yr 4 Mth

N/A N/A 9 Yr 5 Mth N/A N/A 7 Yr 8 Mth N/A 7 Yr 8 Mth

617,632.16 426,768.80 847,200.48 197,200.48 460,393.04 1,504,794.00 460,393.04 1,504,794.00

22.10 19.94 26.20 17.03 9.02 31.00 9.02 31.00

5 Yr 2 Mth 5 Yr 5 Mth 4 Yr 10 Mth 6 Yr 5 Mth 7 Yr 4 Mth 4 Yr 8 Mth 7 Yr 4 Mth 4 Yr 8 Mth

7 Yr 4 Mth 7 Yr 11 Mth 5 Yr 10 Mth 8 Yr 11 Mth N/A 5 Yr 7 Mth N/A 5 Yr 7 Mth

Discount rate 15%. Results from Production and Sale price sensitivity analyses produced identical results.

indicated strong economic viability, with positive NPVs (i.e. IRRs greater than 15%) resulting from most of the sensitivities, and positive IRRs obtained for all sensitivities. Production and sale price ($15/kg) proved to have the greatest affect on profitability (NPV ranging from $460,393 to $1,504,794, Table 9), followed by capital costs (NPV ranging from $197,200 to $847,200, Table 9). Coincidentally, variations in production levels by +/ 20% produced identical results to those witnessed by variations in the sale price from $10 to $15/kg. 4. Discussion 4.1. WTF 20-TPA facility Analysis of the WTF 20-TPA facility clearly highlights that the facility is not forecast to be economically viable. After the 10-year projection period, the facility incurs a large negative net cumulative cash flow of $648,038 and a negative NPV of $707,546 (Table 7). These projections highlight that with such great losses projected it is difficult to accept that any major changes would be likely to develop a viable enterprise at a 20TPA output level.

Previous studies on the economic analysis of commercial RAS facilities of a similar output to the WTF facility clearly highlight facilities that range from being marginally viable to, in some cases, quite lucrative. Of particular interest are the studies by Rawlinson and Forster (2001), Rawlinson (2002), Rawlinson (2004) and Weston et al. (2001). These studies, based on data estimated from industry sources, analysed the profitability of Murray cod cultured in hypothetical, commercial RAS facilities in Australia, with output ranges of between 25 to 30 TPA. Rawlinson and Forster (2001) stated that the 25-TPA RAS facility derived from their model reveals very strong indicators of financial success, with an IRR of 16.7% and NPV of $352,873. These results are in direct contradiction to the results obtained in the present study (NPV$707,546, Table 7), primarily due to the development by Rawlinson and Forster (2001) of a model based on a best-case scenario assuming optimal production and inflated sale price ($20/kg vs. $11 12.50/kg used in the present study). Rawlinson (2002) and Rawlinson (2004) highlighted that the RAS facilities derived from his models at output levels of 25 TPA proved to be marginally viable (IRR

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

325

10.7% and 8% respectively; NPV $99,499 and $114,000 respectively). Surprisingly, Weston et al. (2001) claimed that a payback period of 3 years was achievable from their modelling of a 30-TPA facility. This statement appears highly unlikely and in direct contradiction to the results obtained in the present study, which highlighted that the WTF 20-TPA facility is clearly unviable, with a payback period likely to never be obtained during the project's life. In all cases, inflated sale prices were used for income projections ($1517.50/kg) and data sources were based on industry estimates using best-case scenarios. These comparisons clearly highlight the risks associated with, and question the validity of, developing economic models based on estimated data sources, especially in a developing sector such as the commercial RAS industry. 4.2. Economies of scale Economies of scale were clearly evident as a result of the development of economic models for hypothetical 50- and 100-TPA facilities. In contrast to the nonviability of the WTF 20-TPA facility, the hypothetical 50-TPA facility appeared marginally viable with a positive 10-year net cumulative cash flow of $1,030,300 and negative NPV of $167,651 (based on a 15% required rate of return, Table 7). These projections highlight that whilst the projections of the facility would fall slightly short of the required rate of return of 15% (IRR 11.75%, Table 7), there were large improvements in comparison to the financial performance of the WTF 20-TPA facility. This highlights the importance of economies of scale and the potential for economic viability at higher output levels. The outcomes of the analysis of the hypothetical 100TPA facility further emphasised the importance of economies of scale for the economic viability and potential future success of commercial RAS. These results highlighted that the 100-TPA facility was economically viable and potentially quite lucrative for investors. Results from the economic analysis of this hypothetical facility highlighted a positive 10-year net cumulative cash flow of $3,176,750 and NPVof $522,200 (based on a 15% required rate of return, Table 7). The NPV projections exceeded the required rate of return of 15% (IRR 21.03%, Table 7), highlighting not only further improvements in comparison to the financial performance of the 50-TPA facility, but also a potentially lucrative investment opportunity. Contrary to the large differences in the economic viability of the WTF 20-TPA facility to previous studies (Section 4.1), the limited previous studies on the economics of commercial RAS facilities of similar output to

the hypothetical 50- and 100-TPA facilities generally support the results obtained from the present study. Economic models developed by Rawlinson and Forster (2001), Rawlinson (2002), and Rawlinson (2004) of hypothetical 50-TPA RAS facilities culturing Murray cod highlighted economic viability with IRRs of 18%, 13.9% and 10.3%, and NPVs of $838,000, $415,462 and $377,800 respectively. Noticeably, these reported measures of financial performance were developed by the same author and have decreased over time (perhaps due to the author's ongoing refinement of potential initial overestimations). In comparison, the 50-TPA model generated from the present study produced an IRR of 11.75% and a negative NPV of $167,652 (Table 7), hence indicating only marginal viability. The large range of NPVs between these studies has been predominately caused by the varying discount rates each study has assumed (remembering that a NPVof zero is attained when the IRR equals the discount rate or required rate of return). For example, Rawlinson and Forster (2001) assumed a discount rate of 8%, whilst Rawlinson (2002) assumed a discount rate of 5% (in comparison to the 15% used in the present study). Discount rates of between 5% and 8% (i.e. in this case leading to the portrayal of economically viable facilities) implicitly assume extremely low risk investments. Hence, the projections from these previous studies is questionable, especially due to the high risks associated with investing in commercial RAS, and also considering that these models have assumed best-case production techniques, inflated income sources and sale of all output at the completion of each growout period. Economic models of hypothetical 100-TPA RAS facilities (culturing Murray cod), developed by Rawlinson (2002) and Rawlinson (2004), supported the results from the present study by indicating economic viability at this specific output level. These studies produced IRRs of 16.9% and 13.1%, and NPVs of $1,217,957 and $1,253,641 respectively. Surprisingly, the IRR of 21.03% from the present study exceeded the IRR of these previous studies, hence indicating stronger financial performance than previous models, whilst the NPV was considerably lower at $522,200 (Table 7). As discussed previously, this discrepancy with the NPVs was caused by the differing discount rates used between the studies. 4.3. Key variables The results of the sensitivity analyses from 20-, 50and 100-TPA facilities highlighted that increasing the sale price, maximising production and reducing capital costs were the three key variables which had the greatest

326

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327

effect on the profitability of commercial RAS, hence highlighting key areas where future research and development could have the greatest performance and economic impact. Improvements in these areas will require the constant refinement and development of marketing programs, management techniques, training programs and infrastructure technology. In the case of the 50-TPA and 100-TPA facilities, realistic variations to these parameters proved to be responsible for the difference between viable and nonviable operations. However, as previously discussed in Section 3.4, the large losses projected at the 20-TPA level highlighted that the facility appeared incapable of becoming profitable regardless of any realistic variations to key variables. 4.4. Limitations of the study As with all economic projections, certain assumptions are required in order to create financial models. Hence, the assumptions developed for the present study (Table 3) can be identified as a potential limitation affecting the validity of the results. Nevertheless, in order to improve the potential accuracy of the assumptions, conservative measures were used. However, due to the fact that these assumptions and the overall results are based on one facility only, the results from the present study should be used as a guide. Whilst acknowledging this, it must also be noted that a strength of the present study over previous studies was that it was based on three years of actual commercial data, whereas the majority of previous studies in this field are based on hypothetical data sources, assuming best practice husbandry techniques, optimal production and sale of all output once stock have completed their growout period. This kind of situation is highly unlikely and in many cases has resulted in the development of questionable models based on over-estimated income projections and under-estimated cost projections. Consequently, this may be a direct cause for the many failures in RAS enterprises that have been witnessed in recent times. 5. Conclusion The results from the present study, incorporating real data (for the first time in Australia), have provided vital information regarding the economic viability of commercial RAS. These results have highlighted the major influence of economies of scale, with investor attractiveness not evident until facilities reach approximately 100 TPA, hence establishing what could be

considered to be the minimum output size for an economically viable commercial RAS facility. The contradictory financial performance of the results from the present study to previous studies revealed here clearly highlights the need for more publicly available, commercially based data and further economic analyses of commercial RAS facilities. Perhaps this could be achieved with increased government support (i.e. funding, subsidies, extension services, industry development programs) in conjunction with the ongoing development of commercial scale RAS facilities and research programs. These factors could allow potential investors to appreciate and better understand the consequent risks and uncertainties surrounding commercial RAS, hence potentially reducing the number of future economic failures. In conclusion, whilst this study has highlighted key areas on which to focus future research, it will most probably not be a single measure that will lead to the widespread success of commercial RAS. Rather, the commercial success of this industry will require the need to focus on the combination of all aspects of these types of ventures. Acknowledgements This project was funded by the Victorian State Government, Science and Technology Innovation Initiative. The authors would also like to acknowledge Mrs. Jean Daw, Deakin University Finance Officer and Mr. Peter Kavanagh, Proprietor Warrnambool Trout Farms Pty Ltd for technical assistance throughout the project.

References
Head, W.D., Watanabe, W.O., 1995. Economic analysis of a commercialscale, recirculating, brackish water hatchery for Florida Red Tilapia. J. Appl. Aquac. 5, 123. Honda, H., 1998. Recirculating aquaculture systems in Japan. Proceedings of the 2nd International Conference on Recirculating Aquaculture, Roanoke, Virginia, USA, pp. 124128. Kazmierczak Jr., R.F., Caffey, R.H., 1995. Management ability and the economics of recirculating aquaculture production systems. Mar. Res. Econ. 10, 187209. Larson, K.D., Wild, J.J., Chiappetta, B., 2002. Fundamental Accounting Principles, 16th edition. McGraw-Hill Irwin Publishing, USA. Losordo, T.M., Westerman, P.W., 1994. An analysis of biological, economic and engineering factors affecting the cost of fish production in recirculating aquaculture systems. J. World Aquac. Soc. 25, 193203. Love, G., Langenkamp, D., 2003. Australian aquaculture: industry profiles for selected species. ABARE Research Report, vol. 03.8. Canberra, Australia.

P.N. De Ionno et al. / Aquaculture 259 (2006) 315327 O'Rourke, P.D., 1991. Current status on profits in recirculating systems. Regional Workshop on Commercial Fish Culture Using Water Recirculating Systems, Illinois/Indiana Sea Grant, Illinois State University, Normal, Illinois, pp. 6599. O'Rourke, P.D., 1996. The economics of recirculating aquaculture systems. Proceedings of the 1st International Conference on Recirculating Aquaculture, Roanoke, Virginia, USA, pp. 119. Petty, J.W., Peacock, R., Martin, P., Burrow, M., Keown, A.J., Scott, D.F., Martin, J.D., 1996. Basic Financial Management. Prentice Hall Publishing, Sydney, Australia. Powless, T.F., 1998. The importance of feed to the economic success of the system. Proceedings of the 2nd International Conference on Recirculating Aquaculture, Roanoke, Virginia, USA, pp. 270273. Rawlinson, P., 2002. The economic efficiencies of partial and intensive recirculation aquaculture systems for Murray cod. In: Ingram, B.A. (Ed.), Murray Cod Aquaculture: Now and Into the Future. Victorian Institute of Animal Science, Attwood, Victoria, Australia, pp. 1718. Rawlinson, P., 2004. The economics of partial and intensive Murray cod aquaculture systems. In: Ingram, B.A., De Silva, S.S. (Eds.),

327

Development of Intensive Commercial Aquaculture Production Technology for Murray Cod. Department of Primary Industries Research, Victoria, Australia, pp. 167181. Rawlinson, P., Forster, A., 2001. The economics of recirculation aquaculture. In: Johnston, R.S., Shriver, A.L. (Eds.), Microbehaviour and Macroresults: Proceedings of the Tenth Biennial Conference of the International Institute of Fisheries Economics and Trade. Oregon State University, Corvallis, USA. Staniford, A.J., 1989. The effect of yield and price variability on the economic feasibility of freshwater crayfish Cherax destructor (Decapoda: Parastacidae) production in Australia. Aquaculture 81, 225235. Van Wyk, P., 2000. Economics of shrimp culture in a freshwater recirculating aquaculture system. Proceedings of the 3rd International Conference on Recirculating Aquaculture, Roanoke, Virginia, USA. 6 pp. Weston, L., Hardcastle, S., Davies, L., 2001. Profitability of selected aquaculture species. ABARE Research Report, vol. 01.3. Canberra, Australia.