Supply Chain Analysis of Coconut Biofuel: Findings from

Conan Hales, CFA Consultant to the World Bank

1 Introduction

• Island locations in the Pacific are heavily dependent on fossil fuel imports, which are exposed to increased global oil prices • In 2011, I undertook research in the field of coconut oil used as biofuel as part of a postgraduate thesis for the Imperial College London • This was sponsored by the World Bank and with the co-operation with the department of energy of Fiji – Focus on the pilot operation on the remote island of Koro

2 About the Research

• Purpose of the study is to examine the economic aspect of the entire supply chain for coconut oil biofuel production • Identify how value is shared by the actors involved in the production of biofuel • Research may be used to form policy recommendations • We are looking to answer the question: – Does it make sense to replace diesel fuel with coconut oil?

3 Outline

Context of mini-mills on remote islands Technical feasibility of coconut oil as a fuel Supply-chain analysis Financing the mill Summary

4 Setting the context MINI-MILLS ON REMOTE ISLAND AND GLOBAL PRICES

5 What is a Mini-mill?

6 Where is Koro island?

7 Global Price Comparison of Coconut Oil and Crude Oil

3.00 Approximate time of study is 2.50 when CNO price >> oil price

2.00

1.50 Crude oil

Price (USD/L) Price Coconut oil

1.00

0.50

0.00

1990 1990 1991 1992 1992 1993 1994 1994 1995 1996 1996 1997 1998 1998 1999 2000 2000 2001 2002 2002 2003 2004 2004 2005 2006 2006 2007 2008 2008 2009 2010 2010 2011 2012 2012 2013 Date 8 Question

• Does it make sense to use coconut oil to replace diesel fuel on the remote islands? • Answer depends on: – Total coconut resources available after food consumption needs – The cost of coconut oil production – The landed cost of diesel on the remote island – Non-economic costs or benefits associated with coconut oil production

9 Fuel characteristics and engine modification options FEASIBILITY OF COCONUT OIL (“CNO”) AS A FUEL

10 Technical Feasibility of Coconut oil as a fuel

• Much research published that indicated coconut oil (“CNO”) is a viable alternative to diesel fuel – “Coconut Oil Power Generation: A how-to guide for small stationary engines”, World Bank, 2009

Fuel Specific Density Cetane Kinematic Solidification Flash point energy (kg/m3) number Viscosity Temp. (°C) (°C) (MJ/kg) (cSt 40°C) Diesel fuel 42.6 828 40–55 2–4 -9 >62

CNO 35.8 915 60–70 27 22–25 200–285

11 Modification options of using CNO in diesel engines

• At least 3 options considered for using CNO as a fuel in diesel engines: – Adapting Engines for 80% CNO and 20% Mineral Diesel – Adapting Engines for 100% CNO Use – Biodiesel • First option most suited to remote island setting due to simplicity and cost

Blend: 20% Adapt Engine: Dual fuel system Mineral Diesel

80%

Adapt Engine: Adapt fuel pump, fuel tank, filters,

100% injectors Generator

CNO Engine Compression 100% Biodiesel (Unmodified Engine) Transesterification

Reference: Jan Cloin, “Cocogen: Feasibility Study Into The Use Of Coconut Oil Fuel In Epc Power Generation” 2008 12 Solutions to the challenges of CNO biofuel

Situation Problem that will arise Associated solution High moisture content in copra (greater than • Production rate of CNO drops. • Ensure moisture content in copra is 6%). • Moisture content in biofuel. This has a high decreased by re-drying copra. chance of affecting the power output of engine. Large particles in settling tank entering the CNO • CNO filter clogs and shuts down. • Ensure one or two days are allocated for filter. • Frequent shut downs and washing of filters settling to allow the large particles in the CNO occur delaying oil production time. to settle to the bottom. • Avoid sudden movement of the settling tank or stirring of the settled CNO during decanting for filtering. • Frequent cleaning of the settling tank. CNO in settling tank has high viscosity (or • CNO filter clogs and shuts down. • Use heater installed in the settling tank to solidifies) in the colder months. heat CNO. (Care must be taken not to overheat CNO, which may result in a fire hazard) Large particle sizes in biofuel • The large particles will quickly clog the • Ensure that the CNO filter is frequently generator fuel filters. Clogged fuel filters result cleaned. in reducing the flow of fuel to the combustion • Biofuel blending tanks are covered at all chamber in the biofuel engine resulting in a times. breakdown or ‘choking’. • Avoid any dust/particles entering exposed/open areas in storage and blending tanks. Reference: Singh, R. (2010) Economic Feasibility Of The SOPAC-CATD Biofuel Project, Nadave, Tailevu, Fiji Islands. 13 Sustainability of Feedstock

• Total resource assessment crucial to ensure feedstock sustainability after existing consumption • Satellite imagery technique may be used to count trees – E.g. On island the estimated production is 7.5 million nuts per year using this technique, of which the harvestable number is 5 million and after consumption of 1.5 million leaves 3.5 million for biofuel stock or about 690,000 litres of biofuel. This is about 3x the diesel consumption • Coconut palms typically intercropped with other food sources, reducing competition for land

Reference: Gerhard Zieroth with Leba Gaunavinaka and Wolf Forstreuter “Biofuel from Coconut Resources in Rotuma” 14 The economics of the biofuel SUPPLY CHAIN ANALYSIS OF CNO FUEL PRODUCTION

15 Electricity Generation on a Remote Island

•Crushes •Cultivates •Transports •Expels Farmer •Collects Dryer Mill Generator Electrification •Dries •Filters •Cuts •Settles

16 Overview of the process and prices Process Cost Price Actors Land Coconut • Cultivation palm cc users pc Whole • Collect coconut

Whole • Split Farmers coconut cf

Green copra • Carry pf

Green copra • Dry Copra driers cv • Transport Transportation Dry copra pv

Dry copra • Crusher

Dry copra • Expel

CNO + • Settle solids Decentralised Copra cm Mill CNO + • Filter particles

CNO • Blend

Biofuel • Store pm Distributor Biofuel • Distribution cd pd Energy service • Engine 17 Biofuel C end user Supply chain of Koro Mill

Costs Land user Value of product

Pc = $182 Farmers and electricity Cf = $17+$20 fund

Pf = $657 Village Cv = $71 co-operatives

Pv = $850

Truck Middleman

Decentralised mill in Centralised mill in Cm = $461 Koro Savusavu

Pm1=$715 Pm2=$484 Pm3=$242 Mobile Village electricity Telecommunication committees Operators

Niu Industries Cd = $16,831 18 Pd = $17,545 Share of estimated profits (/T of dry copra)

Oil offtaker 12%

Farmers 43%

Mill 31%

Village co-op Fund 12% 2% 19 Estimated Market Share of Biofuel End-users

Telecommunication 11%

Koro villages 30%

Excess CNO 59%

20 Cost of production of CNO fuel FINANCIAL ANALYSIS OF THE MINI- MILL

21 Economics of the decentralised mill

Copra (kg) Diesel (L) 70% Bf (L) 80% Bf (L) CNO (L) Meal (kg) Costs Rebate Production (CNO vol) 74 176 397 Input 1,000 76 Daily production 105 220 397 350 Price $850.00 $2.38 $2.30 $2.20 $1.80 $0.52 Wages $148.75 Catalyst $40.73 Overheads $69.72 Truck $32.17 0.13c on diesel (/L) $9.82 Subtotal -$850 -$180 $242 $484 $715 $182 -$291 $10

Profit/Loss $311 /day or $6,738 /month

Cost of production (/L) $2.03 $1.96 $1.70 Implied CNO price (/L) $2.27 $2.16 22 Sensitivity of Mill’s Profit to Diesel and Copra Prices

800 700-800

700 600-700

500-600 600

400-500 500 300-400 400 200-300 300 100-200 200 0-100 3.81

Mill's net margin ($/Tdce) margin net Mill's 100 3.14 2.48 -100-0 0

1.81

-200--100

17

1.14

272

374

476

-100 578

680

0.48

782 884

986 Mineral diesel

1088 1190

1292 price ($/L) 1394

-200 1496

Price of dried copra paid to the village co-ops ($/Tdce) 23 Return on Equity for Mill Financing

60%

50%

50%-60%

40% 40%-50% 30%-40% 30% 20%-30%

10%-20% Equity holding 20% 0%-10% Term (yr) 20% 30% 40% 15 -10%-0% 5 -42% -18% -10% 10% 14 13 -20%--10% 6 0% 3% 5% 12 11 7 18% 15% 14% 0% 10

Internal Rate Rate Internal of Return of equity 9 8 29% 23% 20% 20% 8 7 9 37% 28% 24% -10% 30% 6 40%5 Term of lending (years) 10 42% 32% 27% 11 46% 35% 29% -20% 12 50% 37% 31% 13 52% 39% 32% Equity holding (%age of total capital) 14 54% 40% 33% 15 56% 41% 34% IRRs for various tenors and leverage were calculated assuming debt can be financed at 9% All the profit goes to equity as excess cash-flow after paying interest on debt 24 Supply-chain analysis of larger mills

• Centralised milling operations exist in some Pacific island countries (E.g. Fiji, Vanuatu) • Centralised mills benefit from economies of scale for CNO production, however, fuel prices may be expensive if cost of delivery is expensive to remote locations. • Needs further investigation on a case-by-case basis

25 SUMMARY OF FINDINGS

26 Summary of Findings

• Remoteness of the island is a key factor for high diesel prices and lower costs of production • Farmers are significant beneficiaries in the supply chain • Koro mill should be profitable, and should benefit the island’s economy • Access to the outside market for the coconut oil is important • The current distributor is exposed to falls in diesel price, so it is important to find a more reliable oil off-taker • New mills on other islands could be financed on commercial terms • Regular quality control of the biofuel is important for consumer confidence and avoiding long-term technical issues • There is potential to improve traditional copra methods in the supply chain – E.g. Use of solar drying instead of thermal drying which can contaminate the copra – Poor handling of the wet copra increases aflatoxins and FFAs • Once a mill is established, there are potential upgrade opportunities for production of higher quality oil for human consumption and use of other parts of the nut such as the husk and shell 27 Findings from Other Case Studies in the Pacific Islands

There is a consensus that CNO used as a fuel is technically feasible based upon , , Koro, EPC Samoa, ENERCAL Ouvéa, Vanuatu island fuel and PNG Fisheries case studies. Despite the technical feasibility, the four categories of issues which have either prevented or caused pilot projects to fail are: Policy • Subsidies are preventing economic biofuel production in Kiribati and Rotuma • Laws are preventing sale of biofuel for vehicles in Vanuatu Technical • Non-trivial techniques of refining fuel for transport engines in PNG fisheries and Vanuatu cases. • There was no engineer to fix the generator in Welagi, but that was subsequent to the decision to run the generator on mineral diesel Feedstock • Competition from dalo was the primary issue in Welagi • Supply and quality issues existed in Ouvéa • There is also a potential feedstock risk from cyclones Socio-economic • There was a failure to understand the community structure in Vanua Balavu, and farmers sold copra for a better price to the central mill as they did not receive benefit from the local mill 28 Benefits of CNO biofuel

• Potential to reduce fuel import bills • Reduces total greenhouse gas emission balance if it displaces mineral diesel • Compared to solar PV, the levelised cost is cheaper – F$0.73/kWh versus F$0.86/kWh • Relies on basic, proven and affordable technology • Remote islands benefit from energy security by relying less on imports

29 Reference

• The research is publically available on the Linked- in profile under Conan Hales

31