WOOD GASIFICATION ENERGY MICRO-GENERATION SYSTEM IN
BRAZIL- A MONTE CARLO VIABILITY SIMULATION
Glaucia Aparecida Prates
UNESP- Sao Paulo State Univerity, Campus Of Itapeva, Brazil
E-mail: glaucia@itapeva.unesp.br
Erich Gomes Schaitza
EMBRAPA- Floresta- Colombo, Brazil
E-mail: gprates@hotmail.com
Submission: 30/05/2017
Accept: 02/07/2017
ABSTRACT
The penetration of renewable energy
into the electricity supply in Brazil is high, one of the highest in the World.
Centralized hydroelectric generation is the main source of energy, followed by
biomass and wind. Surprisingly, mini and micro-generation are negligible, with
less than 2,000 connections to the national grid. In 2015, a new regulatory
framework was put in place to change this situation. In the agricultural
sector, the framework was complemented by the offer of low interest rate loans
to in-farm renewable generation. Brazil proposed to more than double its area
of planted forests as part of its INDC- Intended Nationally Determined
Contributions to the UNFCCC-U.N. Framework Convention on Climate Change
(UNFCCC). This is an ambitious target which will be achieved only if forests
are attractive to farmers. Therefore, this paper analyses whether planting
forests for in-farm energy generation with a with a woodchip gasifier is
economically viable for microgeneration under the new framework and at if they
could be an economic driver for forest plantation. At first, a static case was
analyzed with data from Eucalyptus plantations in five farms. Then, a broader
analysis developed with the use of Monte Carlo technique. Planting short
rotation forests to generate energy could be a viable alternative and the low
interest loans contribute to that. There are some barriers to such systems such
as the inexistence of a mature market for small scale equipment and of a
reference network of good practices and examples.
Keywords: Biomass,
distributed generation, small-scale, Monte Carlo
1. INTRODUCTION
In the early 80´s, Alvin Toffler, an
American philosopher and futurologist created the term “prosumer” to describe a
mass movement of people who would participate in the production process, both
for the pleasure of building their own products or as part of their economic
life (TOFLER; ALVIN, 1981).
At that time, the “do it yourself”
movement and personal computers were gaining momentum. For (KOTLER, 1986) came
back to the same theme and discussed the implications of the prosumer to
marketing. He classified prosumers in two main groups “The avid Hobbyist”,
including those who spend most of their free time doing something else in an
intense way, and “The Arch prosumer”, those who want to be closer to nature and
do things by themselves and whose motto was “small is beautiful”.
They had no idea that years later
their concept would apply perfectly to the electric power sector and impact the
way electricity generation and distribution were organized. With the
development of distributed renewable energy, especially solar photovoltaics,
and the possibility of connecting generators to the grid, millions of people
started to use and sell energy. This new paradigm changed completely cost
relations, presented challenges to grid management and opportunities for new
innovative businesses (PARAGAND; SOVACOOL, 2016).
The expansion of prosumers and
distributed renewable energy is not homogeneous in the world, due to different
conditions of countries in the world (CRIEKEMANS, 2012).
BRAZIL has 4.3 million family run
small-scale rural properties (family farmers). They represent 84% of rural properties
in number and cover 25% of Brazilian farm land. Property size varies between 1
ha up to 100 ha (MDA, 2016) and a good proportion of the reforestation effort
to achieve the INDC reforestation target is expected to be in these areas.
A concessional credit line with
interest rates of 2.5% and 12-year pay-back, exclusive to family farmers and
renewable energy generation, was created by the National Program of Family
Farming Promotion (PRONAF).
This is the best financial loan
available in the country for electricity generation (MDA, 2016) and the key
question is whether it can enable the reforestation of set-aside areas with
short rotation trees and the production of decentralized electricity by family
farmers. In case of a positive answer, funding and need of legal compliance
could work as an induction engine and the Brazilian INDC would progress
simultaneously in two fronts. However, setting up an energy system requires
matching resource availability with suitable technology.
One of the technological
alternatives of small scale processing of forest biomass for power generation
is gasification of wood chips associated with the use of internal combustion
engines, two well-known technologies. At
larger scales, gasification is cost competitive with combustion and is more
efficient) (BRIDGWATER et al ,2012) but various authors
stress that the cost of small scale-plants could be a limiting factor to
commercial operation, with specific capital costs above US$10,000/kW (BOCCI,
2014; LEE, 2013).
On the other hand, combustion
equipment for electricity generation are large and not appropriate to
farm-level production. In the last few years, a series of small-scale
gasifiers with integrated generators and inverters entered the global market (BLANCHARD,
2015) and offer machines in the range of US$ 1500-3000/kW.
The aim of
this paper is to analyses the technical and economic feasibility of a simulated
small-scale grid connected electricity generation system in which small-scale
farmers plant forests and generate electricity through gasification of wood
chips within farm boundaries, in face of the current legislation and
concessional incentives for small in-farm renewable energy generation, with two
levels of analysis:
·
A specific case study with defined
species, forest costs and production was used to determine the role of
different variables and technological barriers of the system and its
components;
·
The risk of adoption of similar concept
in South, Southeast and Central Brazil were production, costs and material
preparation may vary in a stochastic way.
2. MATERIAL AND METHODS
A mix of 3
different commercial clones of Eucalyptus grandis, E. urophylla and E.
urograndis (A08, G100 and I144) was planted in in 5 different small-scale
family neighbouring farms, located at Ajuricaba Condominium, Marechal Cândido
Rondon, Brazil (central coordinate of farms: 24o36´S, 54o08´W). The forests
were planted by farmers and the International Centre on Renewable Energy –
Biogas (CIBIOGAS), a research and development institute associated to Itaipu
Binational (an energy company owned by Brazil and Paraguay), guided the
establishment of forests, accounted costs and measured height and diameter of
all trees at 3 years old in the area.
Using
SisEucalipto simulator, developed by Embrapa (OLIVEIRA et al, 2014) the yield
in volume of plantations was calculated to each farm for two different plant
densities (3333 plants per hectare in a 1x3m spacing and 1667 plants per
hectare in a 3x2 spacing), 2 rotations of 5 and 6 years, 95% of survival of
plant Prices at the local were collected by CIBIOGAS.
Other
prices were supplied upon consultation to the Secretariat of Agriculture of
Paraná, associations of producers, Forest Institute of Paraná, farmers and
agricultural cooperatives.
Simulations
considered Electrical Capacity Factors varying from 0.55 to 0.7 (3942-6132h of
generation per year). As the loan of PRONAF is payable in 12 years, the
analysis was based on a 12-year discounted cash flow, with Net Present Value
(NPV) being used as the dependent variable and measure of potential success of
the business. A sensibility analysis of each of the input variables described
below was also done. s in the first year. Labour was accrued by CIBIOGAS at
US$2.75/hour, the regional value of general services, even when tasks were
carried out by the farmer.
LCOE
(Levelized Cost of Energy) was calculated and compared with photovoltaic energy
using the methodology described in Blanchard (2015) for a period of 20 years.
The Monte
Carlo simulation, a technique that allows analyzing risk in uncertain
environments, without the construction of complexes equations to model a
situation. The method consists of analyzing the effect of simultaneous
variation of input variables in a dependent output variable by random sampling
values within the variability distribution of each input variables with a very
large number of interactions, so that after a distribution of probabilities of
occurrence can be determined.
In fact, it
performs an aggregation of many “what if” simulations and presents the distribution
of all possible results (FRAUNHOEFER, 2013; REES, 2015) interactions were done
with random sampling of electrical capacity factor, cost of feedstock, price of
sale of energy (with an expected value 10% below consumer price from national
utility companies), specific cost of generator and value of extra heat
generated (as a surplus of heat generation, not used to dry chips), the
independent variables which influence net present value (NPV) of the business.
A
triangular distribution was used for all independent variables as their actual
distribution is not known. According to HUANG et al (2015) this is a reasonable
assumption. Minimum, maximum and most probable values for each variable were
estimated as described above.
The
software @Risk 7.50 from Palisade Corporation was used in association with
Microsoft Excel.
3. RESULTS AND DISCUSSION
3.1.
Forest
Production
As the cost
per hectare in all five Ajuricaba Condominium was the same, it is clear that
areas with better growth present a lower cost per ton of wood. Although
small-scale farmers seldom pay land lease, the inclusion of cost of land in
calculations is common and an annual land lease of US$100 was considered as a
cost of opportunity of land. Cost of production of an odt varied from US$8.00
to US$29.00, with an average of US$13.00.
According
to a recent survey, 1 million hectares of forest plantations exist in the State
of Paraná, half in the hands of large forest companies and the other half in
farmers-land HUANG et al (2015). Both methods suggested in this study are
viable, though there are few or no service providers or experience in drying
wood chips in small-scale in Brazil.
The two
methods are: Acquisition of a combined heat module sold by the manufacturer for
US$5,000, with a capacity of 20 kWh when operating at 18kWe, plus US$500 for a
heat exchanger and a fan for air circulation through the pile. The latent heat
of evaporation of water is 686 kWh/t and 0,9kg must be evaporated for every kWh
generated (reducing moisture content from 100% to 25%), the CHP would generate
twice the required heat.
Felling and
stacking the wood 3 months before chipping, so that wood moisture content drops
to 40%, followed by chipping and then drying, according to the suggestion of the
technical assistance of the gasifier manufacturer, by of sun-drying the
feedstock in canvasses, and if further drying is required using a forced-air
flow from the hot air that gets blown out of the radiator on the engine. The
advantage of this method is that, apart from labor costs, no further investment
is required. The disadvantage is that the method is unreliable in rainy seasons
and the feedstock drying process may be disrupted. The cost of drying,
independent of the strategy, was estimated in US$9/odt.
3.2.
The
Integrated System
Integrating
forest production and generation of electricity is a challenge in terms of
dimensions, planning and logistics. As generator, can operate at different
capacity factors (CF) it will demand different quantities of fuel. At a 0.45
CF, the generator would operate 3942 hours and require 85t of wood dry wood per
year. At 0.7, it would operate 6132 hours and require 132t. This increased fuel
consumption has a consequence on the area of forests that will be annually
harvested.
As the
production of feedstock starts 5 years before any increase in capacity factor a
grade of long term planning is involved in the activity. If there is no forest
to supply the generator, it either stops or must be run with wood purchased
from the market at high prices. Even if the forest supply is not a limitation,
as capacity factor increases, the logistics of harvesting, chipping, drying and
storing wood also increases.
Most
probably, a single small farm will not consume all energy generated at this
scale. Therefore, businesses like that would only work with a group of small
farmers consuming the energy or if partners participate with capital and
benefit from reduced energy rates. In that case, the owner of the forest and
the gasifier would work as a small energy company and the off-farm partner as a
client.
3.3.
Economic
and Risk Analysis
The Local Case
Scenario
If an
average farmer at the Ajuricaba Condominium sets a business under the
conditions presented in Table 1, he would have to invest US$ 51,800 and, after
12 years, the NPV of his business would be -US$51,800. Therefore, in this
situation, gasification is not viable.
TABLE 1: Conditions
For The Local Case Average Farmer At The Condominium Ajuricaba
|
Electricity capacity factor |
0.55 |
Specific Capital Cost ($/kWe) |
2,600 |
|
Generator electrical output power (kWe) |
18 |
On-site electricity
value ($/kWh) = grid price |
0.20 |
|
Generator heat output power (kWTh) |
20 |
On-site heat tariff
($/kWh) |
0.1 |
|
Fuel consumption (kg/kWhe) |
Heat utilization factor |
50% |
|
|
Discount rate (%) |
5.0 |
Annual labour cost for operation (US$) |
3,000 |
|
Yearly operational cost of generator (% of cost
of generator) |
5.0 |
Period of business (years) |
12 |
|
Cost of a shed to storage and generator |
5,000 |
Land rent embedded in feedstock cost (US$/year) |
100 |
|
MAI (odt/year) |
26 |
Feedstock cost ($/odt) |
46 |
|
Regional price of wood chips ($/odt) |
70-80 |
Minimum estimated cost of feedstock ($/odt) |
30 |
When a sensibility analysis is
carried out with the main variables considered as seen of Fig. 1, electricity
value is the variable of higher impact, but as the value considered was equal
to grid price and there is no sense in raising it above grid price, it is not
under farmer´s control. The same happens with specific capital cost, which is
set by the market and cannot be changed by the farmer.
Again, the
value considered in the initial analysis was the same as the sales price in US,
which does not consider freight, taxes etc., and actual prices at the Brazilian
market are likely to be higher. Annual payroll and heat value have little
influence on the business, and therefore electrical capacity factor (CF) and
fuel cost are the two variables under control of farmers that could increase
attractively of the business.
At a CF of
0.55, the cost of wood chips would have to be lower than US$ 10/odt to generate
profit, but simulations show that the lowest possible cost of chips would be
US$30/odt. In fact, if everything is kept with values presented at Table 1 and
the CF is increased to 0.62, the NPV turns positive and the business would
become attractive. If the specific cost of the equipment is increased to
US$3,000/kWh, a more likely value of equipment with freight and tax exemption,
a Capacity Factor of 0.70 would be required to turn the investment attractive
with that fuel cost.
The main
incentive given by the Brazilian government to promote the adoption of
renewable energy systems is the PRONAF loan with interest rate of 2.5% per
year, 3 years of grace and 12 years to pay. With the loan, farmers do not have
to disburse money upfront and all investment in the generator is paid in
instalments after the 3rd year. That changes completely the cash flow of the
business (Fig. 1).
At the
initial condition of a capacity factor of 0.55, the generation would pay the
loan and yearly operational profit would be bordering zero after the third
year. With the specific cost of the generator at US$ 3,000, the generator would
have to run at a 0.6 capacity factor to avoid annual operational deficit.
Figure 1:
Sensibility Analysis and Effect of Loan on Current Account
Running an
equipment 60% of the time is equivalent to operate it 14.4 hours a day every
day or 18.8 hours a day if it is operated 6 days a week. Farmers who deal with
livestock are used to work all week and if the gasifier supports such a work
load, gasification is viable with a loan.
In no
condition a gasifier would generate profit if wood chips were bought in the
market, as its average price varies from US$ 70-80, almost double the average
cost estimated for farmers. This difference between cost of feedstock and
market price is due to farmers not having transport costs and not paying
consumer taxes over the feedstock, both costs embedded in the market price.
In the
initial condition of 0.55 capacity factor, yearly consumption of wood is of 104
odt, and a farmer would need to have at least 4 hectares under management for a
5-year rotation, as EMA is a high 26 odt. A total of US$4,500 of labor and land
rent embedded in annual costs of production would not generate financial disbursements
and could be seen as income by the farmer, as follows: US$ 193 of labor and US$
400 of land rent in forest management, US$ 939 of labor in drying and handling
chips and another US$ 3,000 as labor related to running the generator.
The local case
scenario above was near static, developed in an area with good soils and high
productivity, fixed conditions for electricity value, but it allowed the
understanding how different variables are linked to the business.
If
conditions of analysis are broadened to a generic area in South, Southeast and
Central Brazil, a new range of values must be considered for all variables
(Table 2).
TABLE 2: Range Of
Variables Used With The Monte Carlo Technique For A Broad Range Analysis
|
|
Min |
Expected |
Max |
|
Forest production
(US$/odt) |
5 |
16 |
31 |
|
Preparation of
feedstock (US$/odt) |
22 |
33 |
44 |
|
Total cost of
feedstock (US$/odt) |
27 |
49 |
71 |
|
Construction of a
shed for chip drying and storage (US$) |
0 |
5,000 |
6,000 |
|
Specific Capital
Cost (US$) |
2,600 |
3,000 |
3,500 |
|
Electricity
capacity factor |
0.55 |
0.6 |
0.7 |
|
Value of energy
(US$/kWh) |
0.16 |
0.18 |
0.2 |
|
Value of excess
heat generated (US$/kWh) |
0 |
0.1 |
0.2 |
|
Discount rate (%) |
4% |
5% |
6% |
|
Loan (two separate
scenarios) |
No |
|
Yes |
Instead of using
single values and varying one variable at a time, a Monte Carlo simulation was
used. It takes multiple random values of independent variables within specific
variation distribution and ranges. In this process, it may randomly select a
low cost of wood, a low capital cost of the machine, a high capacity factor of
the generator and present a resulting positive net present value, so the
business would be profitable, conversely, it could take a not so favorable
sample and a negative net present value would be achieved.
For
instance, in the capacity factor distribution, 0.55 was considered as the
minimum acceptable value and only 1.3% of the samples assumed this value, while
6.6% of samples assumed 0.6, the value set as the most frequent for capacity
factor. It repeats the process for thousands of times and presents a frequency
histogram with all possible results. In the current case, there was a
convergence in frequency distributions with 10,000 and 50,000 interactions and
graphs (Fig. 2) represent histograms of 10,000 interactions.
Figure 2:
Probability of Profit of Gasification Systems
The same
two scenarios considered with Ajuricaba farms, business with own capital and
with a loan from PRONAF, were then contrasted. As it can be seen at the “NPV –
no loan” frequency histogram of Fig. 2, only 0.7% of all cases would pay back
without a loan, with an expected negative average NPV of -US$ 143,129 ± 879.74
(90% interval of confidence), meaning that the economic risk of running a
gasifier is absolute. The result of the simulation with a loan is opposite,
with 86.2% of probability of profit, with an expected NPV of US$ 54,311.99 ±
811.89 (90% interval of confidence).
Therefore,
the loan can enable the production of renewable energy systems. Again, part of
the costs of the business are not financial and that may enhance attractively.
Farmers
could plant forests in their “set-aside” land and sell wood to the market.
Stumpage prices of wood vary from US$ 15-30/odt, depending on the region of
Brazil and in areas with pulp companies or with high demand for grain drying,
where wood usually have higher prices, gasification may not be interesting at
all. On the other hand, in areas of low demand for wood, electricity generation
is an alternative, as prices are set by the electricity market and not by offer
and demand. Currently, it is mandatory that utility companies accept micro and
mini-generation if technical conditions of connection are met.
3.4.
3.4.
Levelized Cost of Energy and Solar Energy
The PRONAF
loan is directed to any renewable energy and could have been used in a photovoltaic
system. A PV system of 35-40kW have a capital cost of US$ 2,300 (lowest of 3
quotations for a system installed at Ajuricaba).
According
to the Brazilian Atlas of Solar Energy. The annual mean of daily horizontal
global solar irradiation in any region of the Brazil varies from 1500 to 2500
kWh/m2. If installed in an average Brazilian insulation area, of 2,000
kWh/(m²year), it would have a levelized cost of energy (LCOE) of US$ 0.160 for
a 20 year-period, slightly lower than the US$ 0.162 LCOE of a US$ 2,600/kW
gasifier run with a US$ 49/odt fuel at 0.55 capacity factor and slightly higher
than the US$ 0.151 of the gasifier run at 0.6 capacity factor.
Utility
price is US$0.20/kWh, so both technologies could compete with it if price
remain stable. There is already an emerging market for photovoltaic systems in
Brazil, but the capital cost of equipment is still very high if compared with
the US$1500-2000/kW found in equivalent systems in Europe (OLIVEIRA et al,
2014).
One advantage of the PV system is that there is almost no
labor involved in running it, except from periodic cleaning and eventual
maintenance.
The new
regulatory framework of distributed micro- and mini-generation may attract more
Brazilians to produce their own electricity, but at current capital costs of
renewable energy generation systems, neither gasification nor photovoltaic
generation would be viable within a 12-year period without further incentives.
They would be viable at a 20-year horizon, but a large initial investment would
be required to little economic gain.
4. CONCLUSION
The
development of integrated forestry of short rotation forests and in-farm
gasification of wood chips for energy generation is technically and
economically viable if some enabling conditions are present and some barriers
are overcome. The loan granted by PRONAF is one of these enabling conditions.
With a low interest rate and a long payback period, it requires no initial
investment and increase profitability of the business. If not limited to small
farmers, it would certainly bust the adoption of micro and mini-generation in
rural areas and contribute to the Brazilian INDC.
Among the
barriers to the development of small-scale woodchips gasification systems are
the inexistence of current examples of success, the need of development of
appropriate equipment for feedstock preparation and of a value chain for
small-scale renewables, including the production of equipment in Brazil. A
strategy of demonstration of prospects of micro-generation by research and extension
would contribute to overcome some of those barriers.
REFERENCES
BOCCI,
E. et al. (2014) State of art of small scale biomass gasification power
systems: a review of the different typologies. Energy Procedia, n. 45, p.247–256. Available at: http://dx.doi.org/10.1016/j.egypro.2014.01.027.
Access; 10/02/2015
BLANCHARD,
R. (2015) Biomass heating. Domestic Microgeneration: Renewable and
Distributed Energy Technologies, Policies
and Economics, p.77.
BRIDGWATER,
A.V. et al. (2012) A techno-economic
comparison of power production by biomass fast pyrolysis with gasification and
combustion. Available at: www.elsevier.com/locate/rser.Access:
15/05/2015
CRIEKEMANS,
D. (2012) The geopolitics of renewable energy: different or similar to the
geopolitics of conventional energy?. In: ISA Annual Convention, p. 16-19.
FRAUNHOFER,
I. S. E. (2013) Levelized cost of
electricity-renewable energy technologies. Fraunhofer Institute for
Solar Energy Systems ISE, Freiburg, Germany, Tech. Rep.
KOTLER,
P. (1986) The prosumer movement: A new challenge for marketers. NA-Advances in Consumer Research, v. 13.
LEE,
U. (2013) et al., An experimental evaluation of an integrated biomass gasification and
power generation system for distributed power applications. Applied Energy,
101, p. 699–708. Available at: http://dx.doi.org/10.1016/j.apenergy.2012.07.036.Access:
21/03/2015
DESENVOLVIMENTO
AGRÁRIO (2016) Plano de Safra 2015-2016.
http://www.mda.gov.br/sitemda/sites/sitemda/files/user_arquivos_3/ps01.pdf.
Access: 22/03/2016
MINISTÉRIO
DO DESENVOLVIMENTO AGRÁRIO (2016). Programa
Nacional de Fortalecimento da Agricultura Familiar. http://www.mda.gov.br/sitemda/secretaria/saf-creditorural/arquivos-t%C3%A9cnicos-do-plano-safra.
Access: 12/06/2016
OLIVEIRA,
E.B.(2011) Softwares para manejo e análise econômica de plantações florestais.
Documentos, 216, Embrapa Florestas, Colombo.
PARAGAND,
Y.; SOVACOOL, B. K. (2016) Electricity
market design for the prosumer era. Nature
Energy, v. 1, p. 16032.
REES,
M. (2015) Business Risk and Simulation
Modelling in Practice: Using Excel, VBA and@ RISK. John Wiley & Sons.
TOFFLER,
A.; ALVIN, T. (1981) The third
wave. New York: Bantam books.