Electrification of Remote Villages in Assam (India):

Issues and Case Studies

Deb Chattopadhyay
Adjunct Professor
Information Technology and Electrical Engineering
University of Queensland, Brisbane, Australia
[email protected]

Abstract— Rural electrification in India has received last 10 years [2]. The economics of rural electrification and the
considerable attention and funding lately. However, these efforts need for local/distributed renewable generation has long been
are often disorganized, and do not consider the full set of locally recognized since the early nineties [3], and continually argued
available resources. This study presents our analysis for Assam, until today (e.g., Bhardwaj and Tongia [4]). However, apart
a Northeastern state in India, which has more than 2000 (or, from the simple fact that 100% electrification target has not
25%) of un-electrified villages. Our site visits, discussions, yet been met, each policy initiative has been focused on a
research and analysis spanning over a year, show that it is singular activity regardless of the location, composition of
possible to efficiently combine local resources and distribution load and proximity to the grid. There is also poor coordination
network extension using SWER, to electrify many of these
among these initiatives. For instance, the ongoing rural
villages at a fraction of a cost of the current policy measures
using solar home lighting systems. We also emphasise the need
electrification through the Rajiv Gandhi Rural Electrification
to customize the solution for different areas given the variation Program focuses entirely on T&D network, while the Ministry
in local resources among other things. The sample projects we of New and Renewable Energy (MNRE) is conducting a
have selected are currently being considered for pilot projects, massive Solar Home Lighting system (SHS) program in many
to be rolled out throughout the state and possibly other states. cases in the same area. Given the acute shortage of generation
capacity, especially peaking capacity, it is not hard to see why
Index Terms—Rural electrification, cost benefit analysis, solar, some of the T&D focused programs on their own have not
biomass, Single Wire Earth Return. solved the real problem. Although many of the
villages/regions are endowed with excellent local
I. CONTEXT OF RURAL ELECTRIFICATION IN ASSAM hydro/biomass resources, the “monolithic” SHS programmes
Assam in Northeastern India is among the least-electrified have focused solely on micro-scale solar appliances that have
states in the country with 2000+, or 25% of its 8,000 villages achieved very little “electrification” even after providing
still without any form of electricity. Power supply in massive subsidies, up to 90% of the capital cost. While there is
electrified villages is also in dismal state with poor supply indeed some progress with electrification, they have come at a
reliability and 27% of the energy requirement left unmet. higher cost and lower efficiency, than a flexible targeted
Rural household consumption in the state accounted for 2,000 program could achieve.
GWh of electricity in 2010, which is about 30% of total In this context, this paper discusses a framework around
electricity consumption in the state. The lack of electrification which rural electrification issues should be addressed,
and poor supply reliability has rendered an abysmally low per followed by a set of case studies that have been developed as
capita electricity consumption of just 204 kWh pa for the state, part of an ongoing Asian Development Bank project [5]. Some
which is about a quarter of the national average [1]. Assam’s of the findings also dwell on other real-life projects that we
case is representative of a number of other states in India with have undertaken including the Lighting Asia/India project
low/poor electrification of rural areas, including all of the launched by the International Finance Corporation (IFC) [6].
other states in Northeastern India. Our effort has been to develop a methodology that takes a
Naturally, rural electrification in Assam and several other longer term view of the development of the grid, local
states led to a raft of policy initiatives, followed by funding resources and entails flexibility to find the best possible mix of
both nationally and through international funding agencies. resources to deliver electricity at a reasonable cost. We strive
These rural electrification programs have been underway for to show that in Assam, which is relatively poor on solar
several decades now. Target date of 100% electrification of all resources, a combination of biomass, hydro and low-cost
states have repeatedly been set and reset, including those in Single Wire Earth Return (SWER) technology could all be
2009 and 2012. Overall, it is largely agreed that the policy used to articulate an electrification strategy. Our analysis is
initiatives have been ineffectual, despite a steady growth in based on an extensive site visits, discussions with
funding these initiatives at a rate of 8%-10% every year for governmental agencies, funding institutions, and villagers.
 II. METHODOLOGICAL FRAMEWORK hydro. There has been limited effort in harnessing this
A. Techno-economic Analysis potential, and recent policy initiatives for rural electrification
have focused on mostly SHS – both under MNRE as well as
A general limitation of the rural electrification programs – through international funding initiatives (including ADB,
be it T&D or renewable – is that they lack economic
World Bank and IFC). Figure 2 shows the relative cost of solar
discipline. Establishing a rigorous techno-economic discipline
in project selection is important. One also needs to consider with other renewable options in Assam. These cost estimates
not just the immediate relief that the project brings forth, but a are based on Detailed Project Reports obtained from Assam
medium to long term view of the project, to justify significant State Electricity Board, or sourced directly from project
upfront investment that needs to be recovered over developers for actual projects that are in various stages of
medium/long term. Therefore, a project that puts too much development in Assam. As the figure demonstrates, based on
pressure on the available land, or forest resource, and/or has a real-life project costs, biomass/hydro options are expected to
limited span of benefits because the area may eventually be be more competitive in Assam compared to solar. This is
connected to the grid in less than five years, may not be particularly true when we compare the significantly higher
attractive, even if it provides an immediate access to power. If solar insolation in Northern and Western India (e.g., up to 7
the costs are too high and the benefits do not accrue long kWh/sqm/day) compared to Assam (below 5 kWh/sqm/day)
enough, high upfront investment is unlikely to be a financially that would make solar projects in those states a lot more
viable proposition. On the other hand, the short term hardship attractive from an investor perspective.
from lack of electricity limits economic growth of a region.
Apart from any hindrance to commercial activity, the social
costs of not having access to electricity are formidable. We Figure 2 Comparison of levelised costs (Rs/kWh)
70.0
therefore used a techno-economic assessment shown in Figure 62.2
1 that gives due regard to these costs to form the overall 60.0

framework.
Levelised cost (Rs/kWh)
50.0

40.0

Figure 1 Cost benefit analysis framework 30.0 24.8

Cost of 20.0
14.5
electricity Cost of unserved energy
10.0 6.5 4.9 3.8 3.0 2.9
-
Solar Home Solar PV Solar PV Micro hydro Biomass 11 Biomass Mini hydro Biomass
Light (0.037 Plant - with Plant - w/o (40 kW) kW 120 kW (1.6 MW) 250 kW
Renewable project cost kW) battery (<10 battery (<10
kW) kW)

Note: 1 USD = Approx. Rs 50 in 2011/12. Solar Home Lighting system cost

in the figure above assumes a 5-year life span in light of the problems that
Cost of grid supply have been observed with maintenance and replacement of battery. Doubling
the life would reduce the cost around Rs 37 per kWh.

C. Grid connection using SWER

2012 Economic life 2020 Years

We have estimated peak loading in currently un-electrified

The x-axis in Figure 1 shows time sequence and costs would
change with passage of time. Cost of grid supply in the short- villages in Lower, Upper and Central Assam. The analysis of
term is very high because Assam relies on expensive oil- load data shows there are around 300 large villages across
based peaking support. Grid supply cost may come down if these three regions that have a peak load in excess of 50 kW,
that account for 85% of the total un-electrified population
the generation deficit situation in Assam improves over time
(Figure 3).
through interconnection or new local generation (especially
Since many of these villages are within 30 km of the grid,
hydro) projects as is currently planned, and also sufficient
time allowed for SWER/3-phase supply connection. Cost of connecting these larger villages to the grid is a prominent
unserved energy would increase as the opportunity cost of option although difficult terrain and those located on riverine
electricity with higher economic growth increases. islands would continue to be mostly off-grid. Our analysis of
grid connection options suggest that a 3-phase connection is
Renewable costs may increase as cheaper resources are
not necessarily warranted given the characteristics of the load
exhausted over time, albeit that for large-scale solar projects
and availability of local renewable (biomass/hydro/solar). We
may reduce over time as solar panel costs continue to fall.
have considered 33 kV Single Wire Earth Return (SWER)
network for a pilot case study with the 33/19.2 kV Isolating
Transformer located in the existing sub-station. This will help
B. Range of Renewable Options in cost optimisation, voltage regulation and reliability of
power. There may typically be 10-15 numbers of 19.2/0.4 kV
Assam (and most of Northeastern India) is endowed with distribution transformers of assorted size in the villages.
excellent biomass and hydro including small and micro-
 Figure 3 Cumulative distribution of peak load in Assam villages downside – an issue that needs to be tested through the pilot
study.

D. Tailored Solution Packages

Finally, through our review of the situation and in
discussion with the stakeholders, we have developed four
solution packages, or “models” to cover all the villages.
Biomass and hydro and to a lesser extent solar at the right
scale, along with SWER could provide a useful replicable
model in most cases. More importantly, there are possibilities
for these resources to complement each other. Since grid
supply, even if available, is unreliable and of poor quality,
grid-connected local renewable generation can be an asset to
the system. It is important to consider some of the ground
SWER may provide a good compromise between extensive
realities such as the difficulty of managing revenue collection
three-phase high cost connection to customer premises, and
to pay for replacement batteries for solar. We have analysed
relying solely on local renewable energy. In the context of
available solar resource distribution and biomass capabilities
our analysis, SWER would basically alter the cost of supply
to conclude that a combination of these resources, even
from grid (i.e., the redline in Figure 1) – making it a lower cost
without grid supply, can provide power supply to a cluster of
option at an early stage than would otherwise be the case.
household, or a whole village, without requiring any storage,
Apart from a lower capital cost, SWER also offers lower
as shown in Figure 4.
operational and maintenance costs relative to a three-phase
supply and has fewer points of failure owing to its simpler Figure 4 Load/generation profile for off-grid pilot case study
design. On the downside, it has limited capacity and therefore 140
limited applicability. Nevertheless, it is a time-tested
120
technology for entry level technology that fits the Assamese


village context for the following reasons: 100

Low population density areas away from the main grid – 80

kW

importantly, areas that are not expected to grow heavily 60 Biomass


for at least a decade; Solar
40
Preferably feeder length within 10-50 km range,
although there are SWER applications that go well 20


beyond the 50 km length; 0
Relatively low electricity consumption per household – 1 3 5 7 9 11 13 15 17 19 21 23
Hours
typically, any consumption level above 100 kWh per


month per household would push the limit of SWER; Small hydro is an excellent option in Assam with several
Relatively low cost – only about one-third the cost of a hundred MW of potential already identified. There are only
three-phase line. We have calculated and found the four projects implemented to date in reality though.
probable cost of erecting 19.2 kV SWER line will be Nevertheless, data collected for these and other projects in
around Rs 0.41 million or, approximately USD 8,000 pipeline show the delivered cost of electricity is in the range
per km). Actual cost may further come down if we of Rs 3-5 per kWh (6-10 c/kWh), making these economically
assume a greater span length in view of a single running viable without any form of subsidy. Standalone hydro


conductor at a lower voltage level of 19.2 kV.; therefore feature prominently as one of the four models in our
Lower supply quality because power is often carried on pilot projects.
high impedance line for tens, if not hundreds, of


kilometre; E. Grid Parity and Long Run Grid Supply Cost
Diversity-adjusted peak demand below 1 kW for each


connected load; and Although Assam/Northeast faces significant peak/energy
Maximum earth current below 10 amps. shortage at present, one has to keep a longer term view of
power supply. There are, for instance, major hydro and coal-
On the downside, although there has been considerable fired power stations being constructed and cross-border
improvement in the technology, some safety issues have power trades with Bhutan are in the offing. In a separate
continued to be a source of concern. This includes SWER analysis, we have undertaken extensive modelling of the
lines being live for a longer period after a fault caused by Indian/South Asian power grid to develop a distribution of
falling trees. On the balance, the significant advantages, wholesale power prices in Assam in 2016/17 ( Figure 5). In
especially cost advantage, is expected to outweigh the order to adequately capture potential variation in wholesale
prices, the price projection analysis considers 500 different
scenarios/samples that cover uncertain parameters such as example, if there is a cluster of households that
demand, outages etc to produce a distribution of (annual can use a common solar panel and be connected
average) spot prices. On the x-axis, we have shown range of by Low Tension (LT) network, we have opted
wholesale prices. The y-axis shows the “frequency” or how for it that rather than individual very small
many instances of a particular price outcome occur out of 500 panels;
scenarios. The y-axis also plots the cumulative % of price
f. Replicable across several hundred villages – this
outcomes. For instance, there is a 60% chance of prices being
is likely to be the most popular model for off-
below Rs 3.50 per kWh. 80% chance that wholesale prices
grid generation in Northeast; and
will be below Rs 4.50 per kWh, etc. We have used these grid
g. In the longer term, say beyond the next 5 years
supply costs as part of our economic cost/benefit analysis. In
or even before that, it is possible that majority of
the short-term (say within next 5 years), diesel generation
these villages would be grid connected. The
cost of Rs 10 per kWh (or Rs 10,000 per MWh), may be a
renewable facilities should be designed in such
reasonable indicator of opportunity cost. However, if current
a way that these continue to be effective in
planned projects eventuate, grid supply cost boosted by
meeting part of the local demand in a cost-
cheaper hydro resource locally and from Bhutan may push
effective way.
grid supply costs down – an important factor that any
2. Centralised solar PV: covering four Bramhaputra
investment in rural electrification should consider.
riverine island villages in Dhubri district (Lower Assam)
Figure 5 Probability distribution of wholesale prices in Assam in 2016/17
with the following features:
100 120% a. Permanent settlement of a significant population of
90
1,500 (approx) in a relatively small area of 16 Sq
80
100% Km;
b. Relatively good solar insolation within Assam –
70
80% close to 5 kWh/Sq m/day;
60
c. Flat terrain;
Frequency

50 60%
Frequency
d. Difficult/expensive to connect to the main grid; and
40
Cumulative % e. Replicable across a number of other similar sized
40%
30 island villages.
20
20%
3. Centralised solar PV and grid connection using
10 SWER/3-phase supply: The cluster of villages around
0 0%
Dhubri district in Assam West (Lower Assam), and
Dhemaji district in Upper Assam are good representative
Annual Average Wholesale Price (Rs/MWh) projects with the following features:
a. Reasonably large population to provide good scale
for centralised solar PV. There is a cluster with
population of over 15,000 people of medium to
III. CASE STUDIES FOR ASSAM high density (within 25 square km);
b. Relatively good solar insolation – close to 5
This section summarises the findings of our analysis kWh/sqm/day;
developed for the Assamese villages based on the c. Flat terrain;
methodology discussed in the preceding section. We have d. Fringe of grid with the possibility of SWER or 3-
come up with the following four models: phase supply to:
1. Off-grid solar and biomass combination: Combination i. Complement solar generation,
of solar PV and biomass in Sonitpur (Central Assam) and ii. Meet load growth,
Lakhimpur (Upper Assam) districts with the following iii. Render greater reliability, and
features: e. Replicable for other such large villages or clusters
a. Good quality biomass including crop residues that in fringe areas – the total number may be relatively
can ensure sustained biomass production; small but each such cluster would cover several
b. Solar PV to augment biomass; thousand residential and small commercial
c. Potentially most cost-effective way of producing customers (rice mills, etc).
off-grid power, but also has significant issues 4. Micro/mini hydro with or without grid connection: for
around capacity building/training for O&M and the Karbi Anglong hilly region among others that offers a
financial management of the facility; vast array of hydro development opportunities from a
d. Off-grid facility because of the difficulty in getting few kW up to 20 MW. Depending upon the size of hydro
network access through forest areas and the cost of potential and local population size/density, these projects
distribution network for low density population; may be considered for off-grid or grid-connected
e. Nevertheless, the size of solar panel is selected development. The salient features are:
to make it as cost-effective as possible. For
 a. Potentially cost-effective solution with levelised financial analysis does not assume any subsidy component,
cost of generation in the range of Rs 3-5 per kWh; the findings suggest these projects can be economically
b. Low density and low income population; viable even when these villages eventually become grid-
c. Potential for export of significant surplus amount connected. These findings are important for a replicable
to the grid; and business model for rural electrification. Once the policy
d. Both off-grid and grid-connected models may be framework is revisited to embrace the need for flexibility and
replicable across dozens of sites. In Karbi Anglong optimising locally available resources with grid, we envisage
alone, there are 44 sites identified with a total that rural electrification can be a commercial reality in India.
potential of 73 MW.

Table 1 Summary of selected locations

IV. CONCLUDING REMARKS Location and Size BCR and
Model Characteristics (kW) Cost IRR
Off-grid Typical 300 170 Rs 17.8m, BCR of
Rural electrification in India has progressed rather slowly
solar- household village kW or 4.3
due to lack of flexibility in related policy settings. The biomass may be supplied USD IRR of
biggest handicap in the policy is lack of an inclusive using 50 kWp solar 356,000 10%
consideration of local renewable resources. Assam is among panel (without
battery ) and 120 kW
the lowest electrified states, that has excellent biomass and
biomass
hydro potential – yet, a heavily subsidised solar home Centralised Typical 250 100 Rs 27.5m BCR of
lighting system was thrust upon the state as part of a solar PV household village kWp , or USD 1.25
monolithic National Solar Mission. There is also a massive plant may be partially 550,000 IRR of
T&D programme that is going on for years that are in theory supplied with power 8%
using 100 kWp solar
electrifying villages but with extremely poor reliability of panel (with battery
supply. The two extremes of micro solar appliance and T&D storage). Suitable for
focused programmes have not delivered the desired areas without
outcomes, and have chewed up significant budget over the significant biomass
potential
last decade. We had set out an initiative to look for tailored Solar PV A large cluster of 200 Rs 56.5m, BCR of
intermediate solution that takes advantage of local renewable and SWER village on grid fringe kWp or 4.28
resources, as well as grid connection using SWER can use a solar + USD 1.13 IRR of
technology. We are also mindful of the economic life of any combination of solar 33/19.2 m 16%
panel (200 kWp with kV
renewable development in the event grid supply costs come battery storage) and SWER
down over the years, as is predicted to be the case. SWER for grid
Through an extensive survey of existing programmes and supply
projects, site visits and consultation with stakeholders, we
Micro/mini There are 25 projects 500- Rs 170m, BCR of
devised four broad models of grid supply that adequately hydro below 1 MW in 3000 or USD 6.39
cover about 85% of the un-electrified population in the state. Karbi Anglong and kW 3.4 m IRR of
These models/schemes involve efficiently matching available another 17 above 1 13%
hydro, biomass and solar, together with low-cost SWER- MW.
based connection. Table 1 summarises the key details for
these models. In particular, the last column includes Benefit-
Cost Ratio (BCR) and Internal Rate of Return (IRR) for the V. REFERENCES
project. Benefits for these proposed pilot projects for each of
the four models are calculated using a mix of opportunity cost [1] Central Electricity Authority, Report on 18th Electric Power Survey of
of alternative energy and cost of unserved energy, depending India, 2011, Ministry of Power, New Delhi, India.
[2] S. Krishnaswamy, Shifting of Goal Posts: Rural electrification in India,
on the context (i.e., off-grid or grid-connected). IRR 2010. Vasudha Foundation, Bangalore, March 2010. (Available online:
represents the rate of return at which the (net present value) http://www.christianaid.org.uk/images/shifting-goal-posts.pdf)
of costs and benefits are equal. BCRs for all four models are [3] D. Siyambalapitia et al, Evaluation of grid connected rural
> 1 which renders them attractive and with the exception of electrification projects in developing countries, IEEE Trans on Power
Systems, 1991.
the Centralised Solar PV Plant, they are all significantly [4] A. Bhardwaj and R. Tongia, Distributed Power Generation in Rural
higher than 1, which again highlights the importance of India – A Case Study, Carnegie Mellon University, 2010.
mixing local resources. IRR values, calculated using a [5] D. Chattopadhyay and S. Bose, Assessment of Renewable Opportunities
detailed financial model, are also quite healthy for all in Off-grid Locations in Assam, TA 7378 (IND), submitted to Asian
Development Bank, February, 2012.
projects, although the different life, upfront costs, etc mean, [6] International Finance Corporation, Lighting Asia: Solar Off-Grid
these do not necessarily align with the BCR. Since our Lighting, May, 2012.