Cluster Approach for Effective Decentralization in Off-grid Energy Project: A Case Study from Dhenkenal District, Odisha

Paper Code: 215

Debajit Palit, K Rahul Sharma, Sudhakar SundarayThe Energy and Resources Institute, New Delhi

IV th International Conference on Advances in Energy ResearchIndian Institute of Technology Bombay, Mumbai

12th December 2013

Contents

1. Introduction and Rationale of Study2. The OASYS project in Dhenkanal, Odisha3. Study Objectives4. Methodology5. Results and Discussions

Introduction and Rationale• Decentralised renewable energy power plants have emerged as one

possible solution to providing energy access to remote communities.

• A large number of programmes have been initiated, with varying degrees of success.

• The design approach including the selection of services to be provided, technology, local institution and business models vary from model to model.

• The objective here is to present a design approach for solar mini-grids, adopted under the OASYS project in Dhenkanal, Odisha: with a focus on the clustering approach – technical versus institutional clustering

The OASYS project in Dhenkanal, Odisha

• A multi-consortium project: ‘Decentralised off-grid electricity generation in developing countries: Business models for off-grid electricity supply’, known as ‘OASYS South Asia Project

• Project aims to develop innovative and participatory business models for decentralised off-grid electricity supply in South Asia

• One component of the project is to develop an off-grid delivery model framework and implementation of demonstration project(s), covering un-electrified villages, to test the framework.

• The project area under consideration comprises of four villages and one hamlet, namely Rajanga, Kanka, Chadoi, Baguli and Rajanga Hamlet.

The OASYS project in Dhenkanal, Odisha

• The village cluster is at a distance of 120 km from the state capital of Bhubaneswar.

• The households are mostly from the tribal communities and are completely un-electrified.

• These villages have also not been considered under the RGGVY scheme have also little chance of getting access to electricity through grid extension as they lie inside the Kandhara Reserve Forest which falls under an elephant corridor in the district.

Name of Village Rajanga Kanaka Baguli Chaddoi Rajanga Hamlet(PuranaSahi)

Latitude/Longitude

N 20o34’07.6”E 85o16’26.3”

N 20o32’45.6”E 85o16’0.67”

N 20o33’19.6”E 85o17’47.2”

N 20o32’51.9” E 85o16’38.6”

N 20o34’26.4”E 85o16’24.7”

Total households 34 43 35 12 12

Total Population 140 189 142 46 38

Study Objectives• Clustering (along with distribution lines to clustered villages) is the preferred

approach and prescribed by the DDG guidelines.

• Clustering has benefits due to economies of scale

• However, it is important to assess the benefits of clustering, specific to the sites under consideration – from design, investment, O&M and institutional perspectives.

• Additionally, clustering methods need to be modified for different technologies – solar v/s biomass for instance

• Hence this study was undertaken to compare the techno-economics and institutional benefits of different models of clustering – one which is technical and another which is institutional

Methodology• An energy demand and livelihood assessment exercise has already been carried out using

(Participatory Rural Appraisal) PRA techniques, the results of which are inputs to this paper

• From the resource point of view, biomass, solar, wind and small hydro were considered and solar emerged as the most feasible option. Biomass, though available, is largely in the reserve forest.

• Components of the project as a whole include: solar PV power plants, civil construction, public distribution network and household wiring, institutional overheads, human resources and O&M

• Costs of components are based on quotations taken from local vendors

• O&M costs at 0.5% of capex have been considered for 5 years assuming an inflation of 10%

• 2 cases have been considered:– Distributed power plants in all 5 sites– A centralised power plant at Rajanga (most accessible site)

MethodologyAppliances Capacity Purpose

Grinder 2 HP at Rajanga For grinding turmeric and chilli powder

Electronic Weighing Scale and Sealing Machine

Weighing Scale – 10-20 W at RajangaSealing Machine – 150 W at Rajanga

For accurately weighing and packaging of ground turmeric and chilli powder. These have wider application since users can use the ma-chine for other products as well and bring about standardization through accurate weighing, delinking them from the middle-men

Saal Leaf Plate Pressing Ma-chine

0.5 HP at Rajanga The community is currently engaged in the stitching of basic Sal leaf plates, called Khalis. By pressing two Khalis together in the machine, a firm plate can be moulded which has a much higher value in the market.

Water pumps

2 HP each at Rajanga and Kanaka

Agricultural is possible only during the monsoon and the farm lands lie unused during the rest of the year. Hence, to improve agricultural yield, especially of high value crops such as brinjal, water pumps are being provided to initiate activities during the non-monsoon period.

Water Purifier 50 LPH, 75 W each at Rajanga, Kanaka and Baguli

For provision of clean drinking water

In addition to the above, provision for the following services has been included in the sizing of the power plant: TV-DVD at the community centres (for educational and training purposes) Community centre lighting and fans (to create a resource/community centre for people to work and also for meetings,

discussions, trainings, etc.) 2 light points and a plug point at every household

Methodology

Case 1- Based on the load profiles and one day of autonomy for night time loads, the total PV

and battery size have been calculated

- 5 independent systems have been designed

- The three power plants catering to a larger number of households, community, productive and/or agricultural loads (in Rajanga, Kanaka and Baguli) are designed as AC power plants. Taking costs into consideration, the two smallest sites (with 12-15 households with basic 2 light points and mobile charging requirements) are being provided with DC micro grids. However the quality of service (lumen output of lighting and hours of operation) are the same in AC or DC

- The pumps have been connected to the main power plants (rather than an independent solar PV set) via a distribution line (at a slightly higher cost):- (i) cater to any future requirements of increase in pump size and - (ii) cater to inrush current required by the pump when it is switched on.

- Demand side management (alternating pump with other livelihoods) has been factored in to optimise PLF

Site AC/DC Plant Design Length of the distribu-tion line along the

longest feeder (met-ers)

Total length of the distribution

line (meters)

Rajanga AC 6 kWp SPV power plant, 6 kVA in-verter (grid tied), 48 V 500 Ah bat-tery bank

1500 2100

Kanaka AC 5 kWp SPV power plant, 5 kVA in-verter, 48 V 600 Ah battery bank

900 1400

Chadoi AC 2.5 kWp SPV power plant, 2 kVA inverter, 48 V 500 Ah battery bank

700 1100

Baguli DC 200 Wp, 24 V 100 Ah battery bank, 15 A charge controller

200 300

Rajanga Hamlet

DC 200 Wp, 24 V 100 Ah battery bank, 15 A charge controller

200 300

MethodologyAlternative Designs: Cases 2 and 3

Case 1 was arrived at with the objective of optimizing system design, reducing losses and minimizing projects costs

In cases 2 and 3 we consider the alternative of a central power plant at Rajanga since it is the easiest to access and leads to lowered transport costs and reduction in delays

MethodologyCase 2: Increased transmission infrastructure but no change in the cumulative size of the centralised power plantEconomic Voltage = 5.5 +)L is the length of distribution line in km and kVA is the Power per phase required to be transmitted

Case 3: No change in the transmission infrastructure, but increase in PV system size to compensate for losses accruing from extending the existing single phase transmission network. - Single phase extension considered- Two options: extending line using wire gauge used in local PDN and extending the

line using wire of 45.6 mm2.

- Other cable options can also be considered, however these two options were selected taking cost into consideration.

- Two inverters of 9 kVA each have been considered, one to cater to the day load and another for the night load, so that the inverters can work at the rated capacity for higher efficiency.

Results and Discussion

Case 2:

Village/ Hamlet

Economic Voltage (kV)

as per calculationas per allowable stan-dard voltages

Rajanga hamlet 3.4 3.3Kanaka 7.3 6.6Chadoi 6.7 6.6Baguli 7.3 6.6

As expected, the design becomes unfeasible

Cost of transmission cables, step-up and step-down transformers and protection devices would be several times of the power plant cost

Results and DiscussionCase 3- High power loss unto 11% in one case occurs- To compensate for this loss: For option -1, where same local cables are extended,

the size will increase by 1,660 Wp and for option - 2, where a 45.6 mm2 cable is used, the power plant size would need to be increased by 830 WpSite AC/DC Plant Design Length of the distribution

line along the longest feeder (meters)

Total length of the distribution line

(meters)Rajanga AC 14.6 kWp SPV power plant, 2*9

kVA inverter (one grid tied), 96 V 800 Ah battery bank

3,700 13,800

ParticularsCost (INR)

Case 1 Case 3Solar Power Plant 3,162,229 2,926,000Distribution Line 1,748,000 5,244,000Household connections 272,000 272,000Construction Requirements 1,983,923 1,523,861Annualised O&M cost over a 5 year period 241,894 100,948TOTAL 7,166,152 9,965,861

Results and Discussion- 40% higher cost of implementation with O&M costs factored in for a 5 year period

- Transmission costs however have increased 3 times

- There may be some operational benefits from a centralised system, however a strong case can be made for distributed systems (with a single entity to manage the cluster) owning to increased capex and overall reduction in system efficiency. Also for such a community, maintaining the health of such a long PDN will be difficult.

- This may be the case for solar (which is highly modular) but not for biomass or other RE where there is a limit on the minimum capacity of power plant and a clustering of loads around a single power plant is preferred

- Population density is another factor where results will vary, owing to the sparse population of the area under consideration

- Future research can look into similar techno-economic analyses of implementation costs and can have useful implications for programmatic implementation of renewable energy technologies.

Thank youk.sharma@teri.res.in