Optimization of decentralized energy systems using biomass resources for rural electrification
in developing countries
Diego Silva, Toshihiko Nakata
Tohoku University, Graduate School of EngineeringDepartment of Management Science and Technology
Sendai, Japan
June 23, 2009
IAEE 32nd International Conference, San Francisco
Contents
• Introduction• Methodology• Results and discussion• Conclusion and future tasks
2Diego Silva and Toshihiko Nakata - Tohoku University
• Introduction• Methodology• Results and discussion• Conclusion and future tasks
3Diego Silva and Toshihiko Nakata - Tohoku University
Energy access and the MDGs
4Diego Silva and Toshihiko Nakata - Tohoku University
1.6 billion without electricity in 2005 (IEA)
2.4 billion rely on traditional biomass in 2005 (IEA)
1.6 million deaths due to indoor air pollution in 2004 (WHO)
Electrification schemes
5Diego Silva and Toshihiko Nakata ‐
Tohoku University
Grid electricity
Biomass
Other renewables
Extension of gridRural areas
Urban area
Areas connected to the grid
Remote areas outside the grid
Fossil fuels
Foreign resources
Local resources
Centralized
Decentralized
Previous studies ‐
Rural electrification
• Evaluation of technologies or a set of technologies.
• Decentralized electrification with renewables or combined
with diesel (hybrid configurations).
• Allocation of agricultural resources for energy.
• Optimization methodology is most common approach.
6Diego Silva and Toshihiko Nakata ‐
Tohoku University
Research goal
• Evaluate a decentralized energy system for rural
electrification in developing countries using local biomass
resources.
• Incorporate into the optimization differences in energy
consumption and income levels between urban and rural
areas.
7Diego Silva and Toshihiko Nakata ‐
Tohoku University
• Introduction• Methodology• Results and discussion• Conclusion and future tasks
8Diego Silva and Toshihiko Nakata - Tohoku University
Target area• South America, Colombia, Meta department
• Proximity to capital city
• Agricultural activities – Rice, sugarcane, oil palm
• Areas connected to the
electricity grid– National Interconnected System (NIS)
• Areas not connected to the
electricity grid– Non Interconnected zones (NIZ)
9Diego Silva and Toshihiko Nakata ‐
Tohoku University
Target area• Energy supply expensive in remote areas (NIZ)
– Electricity price is same as paid by middle income houses in the
interconnected area
– Diesel fuel 1.8 times more expensive
10Diego Silva and Toshihiko Nakata ‐
Tohoku University
NIS
Total population 724,929
Population with electricity 86%
Population without electricity 14%
NIZ
Total population 18,668
Population with electricity 86%
Population without electricity 14%
Proposed energy system
11Diego Silva and Toshihiko Nakata ‐
Tohoku University
-Energy resources Energy demandEnergy transmission
and distributionEnergy conversion
technologies
Foreign resources Residential sector
Local resources
Electric grid
Diesel fuel
Rice husk
Bagasse
Sugarcanewaste
Dieselgeneration
Directcombustion
Gasification
Electricitysupply NIS
Electricitysupply NIZ
Pyrolysis
NIS-Urban
NIS-Rural
NIS-Populatedcenter
NIZ
Forest biomass
NIS: National interconnected systemNIZ: Non interconnected zones
Results
Linear programming• Minimumoverrun (net)costs
Sensitivity analysis • Biomass share
Optimization model
Flow of analysis
12Diego Silva and Toshihiko Nakata - Tohoku University
Parameters • Costs of technologies and resources
•Conversion efficiency
Constraints• Energy resources stock• Electricity demand• Resource location• Utilization of agricultural wastes
Additional input• Urban-rural differences (prices, resources)
• Socio-economic strata (prices)
Energy system structure• Technologies• Resource
allocation
Energy system performance • Costs• CO2 emissions• Regional
differences
Input data
Optimization• Linear programming (LP) formulation
13Diego Silva and Toshihiko Nakata ‐
Tohoku University
( _ cos ) ( _ ) klk l
B Tota l ts To ta l revenue b= − = ∑ ∑
ijk j ijkj i
k l kl kl klkl
l
c qd b p d
d
η⎡ ⎤⎢ ⎥ × − =⎢ ⎥⎢ ⎥⎣ ⎦
∑ ∑∑
ijk ij k
q r≤∑ ∑
Minimize overrun costs : Min B
bkl
: overrun costs (net costs)B : total overrun costscijk
: unit electricity generation costdkl
: electricity demandpkl
: unit price of electricityqijk
: primary energy resource ηj
: electricity conversion efficiencyi : energy resourcej
: energy conversion technologyk : locationl : energy demand sector and socio‐economic stratum
Case setting
14Diego Silva and Toshihiko Nakata - Tohoku University
• Baseline
• Cases analyzed– Biomass‐Remote: biomass based electricity for areas outside the grid
– Biomass‐Rural: biomass based electricity for all rural areas
– Biomass‐All: biomass based electricity for all areas
Input data
15Diego Silva and Toshihiko Nakata - Tohoku University
Availability and cost of resourcesResource Stock Cost
ton/yr US$/kgDiesel fuel - 0.290*
Rice husk 43,840 0.080
Bagasse 14,577 0.010
Sugarcane wastes 23,122 0.014
Natural forest wastes 198,158 0.030
Planted forest wastes 138,772 0.030*Diesel fuel cost for NIZ is US$0.530/kgCalculated based on data from UPME (2003).
Input data
16Diego Silva and Toshihiko Nakata - Tohoku University
Features of conversion technologies considered
Conversion technologyEfficiency Capital cost O&M costs
% US$/kW US¢/kWh
Diesel 30 300 1.70
Direct Combustion 17.5 2,300 0.05
Gasification 23.9 4,200 0.07
Pyrolysis 24.7 3,600 0.21Data from Solantausta and Huotari (1999) and UPME (2000)Scale of plants = 2 MWRate of return = 10%, lifetime = 20 years.Baseline for emissions reduction : NIS served with the grid and NIZ with diesel generationCO2 emissions factor for grid electricity and diesel generation are 0.439 kg/kWh and
0.882 kg/kWh respectively.
Input data
17Diego Silva and Toshihiko Nakata - Tohoku University
LocationElectricity demand
GWh/yr
NIS-Urban 192,011
NIS-Rural 28,351
NIS-Populated center 5,984
NIZ 3,517
Demand-Total 228,819
Calculated from data in UPME (2000) and SSP-SUI database.
Electricity price by socio-economic stratum US¢/kWh
1 2 3 4 5 6 Other
3.5 4.2 6.0 7.0 8.4 8.4 8.4
6.0 n.a. n.a. n.a. n.a. n.a. 14.4
Ramirez Gomez, S. (2007).
• Introduction• Methodology• Results and discussion• Conclusion and future tasks
18Diego Silva and Toshihiko Nakata - Tohoku University
0
200
400
600
800
1,000
Grid
_ext
ensi
on
Die
sel_
Gen
Dir_
Com
b
Gas
ifica
tion
Pyr
olys
is
Grid
_ext
ensi
on
Die
sel_
Gen
Dir_
Com
b
Gas
ifica
tion
Pyr
olys
is
Grid
_ext
ensi
on
Die
sel_
Gen
Dir_
Com
b
Gas
ifica
tion
Pyr
olys
is
Grid
_ext
ensi
on
Die
sel_
Gen
Dir_
Com
b
Gas
ifica
tion
Pyr
olys
is
Baseline Biomass-Remote Biomass-Rural Biomass-All
Res
ourc
e al
loca
tion
(103
GJ/
yr)
Planted_forest_wasteNatural_forest_wasteSugarcane_wasteBagasseRice_huskDiesel_fuel_NIZDiesel_fuelElec_Grid
Energy system structure
19Diego Silva and Toshihiko Nakata ‐
Tohoku University
System performance
20Diego Silva and Toshihiko Nakata ‐
Tohoku University
Baseline Biomass- Remote Biomass-Rural Biomass-
All
Overrun costs 103 US$ 3,889 6,539 6,588 11,463
Unit electricity cost UScent/kWh 7.4 8.4 8.5 10.4
Ratio of overrun costs to total costs % 21 31 32 44
Total CO2 emissions 103 t-CO2 117,452 85,211 84,331 0
Unit emissions reduction kg-CO2 /kWh 0.00 0.13 0.13 0.48
0.00
0.05
0.10
0.15
0.20
Baseline Biomass-Remote Biomass-Rural Biomass-All
Uni
t ele
ctric
ity c
ost (
US
$/kW
h)
Urban/rural differences
21Diego Silva and Toshihiko Nakata ‐
Tohoku University
NIS-UrbanNIS-RuralNIS-Pop_CenterNIS-Non_elecNIZ-RuralNIZ-Non_elecAverage
0
2,000
4,000
6,000
8,000
10,000
12,000
Baseline Biomass-Remote Biomass-Rural Biomass-All
Ove
rrun
cost
s (1
0 3
US
$/yr
)
NIZ-Non_elec
NIZ-Rural
NIS-Non_elec
NIS-Pop_Center
NIS-Rural
NIS-Urban
Urban/rural differences
22Diego Silva and Toshihiko Nakata ‐
Tohoku University
-25
0
25
50
75
100
Baseline Biomass-Remote Biomass-Rural Biomass-All
Rat
io o
f ove
rrun
cost
s to
tota
l cos
ts (%
)
NIS-UrbanNIS-RuralNIS-Pop_CenterNIS-Non_elecNIZ-RuralNIZ-Non_elecAverage
Sensitivity analysis
23Diego Silva and Toshihiko Nakata ‐
Tohoku University
0
25,000
50,000
75,000
100,000
125,000
0
10,000
20,000
30,000
0 20 40 60 80 100
CO
2em
issi
ons
(103
t-CO
2/ y
r)
Cos
t (10
3U
S$/
yr)
Share of biomass in electricity generation (%)
Total cost
Overrun cost
CO2 emissions
0
5,000
10,000
15,000
20,000
0 20 40 60 80 100
Cos
t (10
3U
S$/
yr)
Share of biomass in electricity generation (%)
Local resources
Foreign resources
Discussion
• Biomass potential for electricity generation (direct
combustion of biomass).
• Opportunities and barriers for decentralized electrification
with local biomass resources.– Emissions reduction.
– Reduced overrun costs in remote areas.
– Impact on rural (local) development, local business, income and
employment (biomass energy supply chain).
– Balancing costs of foreign resources with use of local resources.– Stable supply of electricity.– Costs, investments, prices (transportation), financial mechanisms
(subsidies).
24Diego Silva and Toshihiko Nakata ‐
Tohoku University
• Introduction• Methodology• Results and discussion• Conclusion and future tasks
25Diego Silva and Toshihiko Nakata - Tohoku University
Conclusion
• Possibility of decentralized electrification with local biomass
resources (direct combustion of agricultural and forest wastes).
• Optimal system for decentralized electrification results in – Electricity cost 8.4‐10.4 UScent/kWh, higher than baseline
(7.4 UScent/kWh).
– Overrun costs of biomass based system 1.5 to 4 times larger.
– Emissions reduction 120x103
t‐CO2
/yr compared to baseline.
• Impact on rural development. – Smaller differences in the proportion of overrun costs.
– Costs of local and foreign resources.
26Diego Silva and Toshihiko Nakata ‐
Tohoku University
Further analysis
• Improvements with respect to factors overlooked by the
model (transportation, stable supply).– Enlarged scope of the energy system (heat demand, other sectors).
– Plant scale considerations.– Transport costs, resource geographical distribution (GIS).
• Alternative model formulation (multi‐objective programming,
dynamic programming).
27Diego Silva and Toshihiko Nakata ‐
Tohoku University
THANK YOU FOR YOUR ATTENTION!
Diego Silva and Toshihiko Nakata ‐
Tohoku University 28