Among the claimants for the title of “world leader” in renewables development in remote areas the island of Eigg (population 90) off the west coast of Scotland, which since 2008 has been obtaining over 80% of its electricity from a custom-designed hybrid system, probably has the best claim. This post reviews operating data that have become available since I posted Eigg, a model for a sustainable energy future in September 2014. It concludes a) that while the project has delivered good results it is inefficient (overall capacity factor 11%), b) that Eigg will probably never be able to do away entirely with diesel backup and c) that the project owes its existence to the fact that 94% of the capital cost was financed by grants. It is economically unviable on a stand-alone basis.
The generation data presented below are from a 2015 paper authored by Schmiel and Bhattacharyya , a 2014 paper entitled Off-grid Electricity System at Isle of Eigg: Analysis and evaluation by green energy technology, authored by Bhattacharyya, plus an article from Eigg Electric giving the local perspective.
These were described in the earlier post referenced above, complete with photographs, and will be only briefly recapitulated here. Details of individual units can be found in the above references:
- Run-of river hydro– Three plants of 10kW, 19kW and 100kW, total 119KW
- Solar PV: Three arrays of 9.9kWp, 21kWp and 22.5kWp, Total 53.4kWp*
- Wind: 4x6kW turbines, total 24kW
- Diesel: 2x80kW generators, total 160kW
- Total renewable capacity: 196.4kW
- Total diesel capacity: 160kW
- Renewable + diesel: 356.4kW
*The project began with the 8kWp solar installation, so at project startup the total installed capacity was 8+119+24+160 =311kW. This number is used in the economic assessment discussed later. The additional 22kWp was added in April 2011 (it was said to be snowing at the time) and the remaining 22.5 kWp was added some time in 2013. The date is not specified but was probably after March.
Additional backup is provided by 60 kW of lead-acid batteries with 3 hours 40 minutes duration at rated output, representing ~220 kWh of storage capacity, sufficient to fill Eigg’s electricity demand for about six hours:
Generation data are summarized in Figures 5 and 11 of Schmiel & Bhattacharyya. Figure 5 plots total monthly diesel and renewable generation from November 2008 through August 2012 and Figure 11 plots diesel and renewable generation broken down by source from March 2012 to March 2013. I converted these into numbers by counting pixels on Microsoft Paint (1 pixel width = 160-170 kWh) and combined them into the November 2008 through March 2013 generation plot shown in Figure 1:
Figure 1: Monthly diesel and combined renewables generation, November 2008 through March 2103.
The increase in the percentage of diesel generation as hydro dries up in the summer months is evident. Also of interest is that in only five of the 53 months (March, October and November, 2009, October and December 2011) was Eigg able to get by with no diesel generation at all. Nevertheless, during the period shown – which covers 4 years and 5 months – renewables filled 84% of island demand, well in excess of comparable projects like King Island (65% projected but apparently yet to be achieved), Gorona del Viento (37% to the end of June 2016) and Gapa Island, Korea (32% after 5 years of operation).
Figure 2 shows the data for the 13 month period between March 2012 and March 13, when renewables generation is broken down by source:
Figure 2: Monthly generation, March 2012 through March 13 with renewables generation segregated by source.
During this 13-month period Eigg supplied 82% of its electricity with renewables – 61% from hydro, 12% from wind and 9% from solar – although hydro generation dried up almost entirely in June and wind and solar could not come close to replacing it. At least some diesel generation was in fact needed to fill demand in all 13 months. Table 1 provides details on actual monthly generation, percentages-of-total and capacity factors (note that figures are approximate):
Regarding capacity factors, the 19% value for hydro suggests that operations are heavily constrained, either because of stream flow restrictions or because production exceeded demand and had to be curtailed, or because of a combination of both. (No data are available on curtailment or on any grid stability problems that may have been experienced.) The 19% for wind seems low and the 10% for solar a little high, but they are both in the general range of what we would expect. Figure 3 isolates the wind and solar components. Wind does not change much from month to month. Solar, however, peaks in May rather than June/July, possibly as a result of cloud cover variations.
Figure 3: Monthly wind and solar generation, March 2012 through March 2013
Table 2 of Schmiel & Bhattacharyya, reproduced below, lists the sources of project funding:
Eigg (Island Trust and residents) paid only 6% of the project’s cost. The remaining 94% came from grants. In fact, after winning a £300,000 share of the National Endowment for the Arts and Sciences Big Green Challenge award in 2010 Eigg has come out ahead on the deal.
I made the following crude estimate of the levelized cost of electricity (LCOE) for the Eigg project using the NREL calculator and the following simplified assumptions:
• Cost/installed kW = £1,664,828/311kW (initial installed capacity) = £5,351
• 25 year life, 5% cost of capital (ref. 1)
• Capacity factor 11%
• O&M costs : Ref 1 gives £30,000/year for full time maintenance. Maintenance is currently part-time so I cut this to £20,000. £20,000/311kW = £64/kW/year.
• Diesel fuel costs: 2kWh/liter, 50,000kWh/year, £1.00/liter = £25,000/year. £25,000/311 = £80kW/year. Added to O&M costs = £144kW/year.
The NREL calculator run is shown in Figure 4 below:
Figure 4: NREL calculator run
The levelized cost comes out at £0.54/kWh, over twice mainland retail rates. But Eigg can still sell its electricity at the old rate of £0.20/kWh because it got the system for free. (The levelized cost falls to approximately £0.14/kWh at a capital cost of zero.)
The variable that has the largest impact on LCOE is the capacity factor. A number of previous posts have pointed out that combining a high percentage of renewables generation with diesel backup reduces the capacity factor to low levels and that this has a strongly negative impact on project economics. As shown in Figure 5, Eigg is an example:
Figure 5: Levelized cost of electricity versus capacity factor, Eigg
Despite its small size the Eigg system contains the basic ingredients common to all high-penetration renewables systems, such as energy storage and a smart grid. It also suffers from the same failings, chief among which are its dependence on weather conditions and inadequate energy storage capacity. The project is over-reliant on hydro generation, which dwindles or sometimes even dies out altogether every summer, and the 220kWh of battery storage is again orders of magnitude too small to allow surplus winter hydro to be stored for summer re-use. The hydro shortfalls could be eliminated by increasing solar and/or wind capacity by a factor or four or five, but this would result in substantial curtailment and reduce the system capacity factor even more. As a result Eigg will probably never be able to get rid of its diesel backup, although with 80-85% renewables penetration this is not a major issue.
What is a major issue is that Eigg, like its sister projects at Gorona del Viento and King Island, requires hefty subsidies to make it pay. Renewable energy will never take off until it can stand economically on its own two feet.
Endnote: Demand Management
I was going to write a section on Eigg’s reported success with demand management until I came across this. According to the Case study summary, Isle of Eigg Heritage Trust average (2010) annual electricity use per household is just 2,160 kWh. But according to the Figure 1 data total generation in 2010 was 306,000 kWh, which for the ~38 occupied households on Eigg at the time works out to about 8,000 kWh/household, almost four times as much. The capacity factors on Table 1 further suggest there is nothing seriously wrong with the generation data. I have emailed Eigg Electric to see if they can shed any light on this discrepancy.