Do Solar Farms Work in Ireland?

2015-04-17

Somebody asked me what I thought about this: the short answer is that at this point solar farms may make sense in Ireland.

NPR had an interesting 3-minute slot last week, reporting on the relative merits of purchasing a domestic solar electric system versus leasing one. The programme focused on two neighbours in New Jersey. One had bought panels, the other had leased them.

The homeowner that purchased the panels benefitted from a capital grant equal to about a third of the cost of the panels in the form of a tax credit. She also receives a Solar Renewable Energy Certificate for each 1,000 kWh of electricity generated by the panels. These can be sold to utilities looking to outsource their renewables obligations for about $200 each. She generates about seven of these per year. With these subsidies, plus the revenue from selling the energy, the array will pay for itself in under ten years. There's a bit of administration involved.

Does it make sense in Ireland?

As always, it depends. As with wind energy, these include the resource at the site, i.e. how sunny it is, the cost and effectiveness of the hardware and the price paid for the energy from the panels, together with other subsidies.

David MacKay outlines how to go about estimating the output from a solar panel in his Sustainable Energy Without the Hot Air, on Pages 38 and 39.

If all the electricity produced by a solar panel can be used to offset electricity purchased from the grid, and assuming that installation and maintenance costs are small, then a panel makes sense, with a pay-back of around five years. Using average sunshine numbers for Ireland, about 2.75 kWh m^-2 day^-1, a small array that offsets some of the base-load requirement (e.g. fridge, appliances on stand-by) of a home in Ireland might make sense right now.

(As a note, this works out as about 3.6 kWh for each 10 m², assuming a conversion efficiency of around 13%, more conservative than David MacKay's estimate of 5 kWh per 10 m² for the UK and a good deal less than the 4.2 kWh per day measured by a Dublin university.)

Most of the electricity is generated when it's not needed, i.e. in the middle of the day. So going beyond base-load, the options are to install relatively complicated and expensive battery systems to store this excess or to sell the energy to the grid.

In Ireland, the price paid by utilities for energy from domestic solar arrays is much lower than the price that home owners pay for grid-supplied electricity. For this reason, if most of the output from an array is sold to the grid rather than used, then it does not make financial sense.

(As a note, the sun's resource changes during the year, i.e. there are far fewer sunshine hours in winter than in summer; the average sunshine hours at a site takes this variation into account. It's worth noting that this is the same problem of energy oversupply when it's not needed and lack when it is needed, just on a longer time-scale. The solution is the same, i.e. grid or battery.)

The price of panels continues to fall rapidly. At present, a figure of around 1.50 euro per watt of solar panel is not a bad approximation, i.e. you can expect to pay around 150 euro for a 100 watt panel. If this price was to fall towards one euro per watt, then solar arrays might make sense, even if most power is sold to the grid.

You can expect to pay at least another 1.50 euro per watt for installation. There's also the inverter required to change DC to AC - these cost a few hundred euro.

Right now, with subsidies along the lines of those in New Jersey, solar would make sense for domestic installations in Ireland; at present, none exist.

How about utility-scale solar power?

On poor farm land, e.g. in rocky regions or on mountainsides, does a large-scale solar installation make sense? We can assume that the infrastructure is already in place, i.e. cables to take the power from the site to the grid and a connection point, i.e. power electronics and a transformer. Building this infrastructure could represent a significant additional cost.

To get an approximation to an answer to this question, I wrote a straightforward Python script. The following assumptions were made in setting up the model:

The simple solar farm model calculates an approximate AEP and number of 100 W panels based on an average insolation across a year (kWh m^-2 day^-1), available area and time period, i.e. usually a year. The farm is assumed to require 1.5 times the area of the panels for fittings, roads and so forth.
The return on investment assumes an approximate panel cost, an approximate installation cost, an average electricity price, a Feed-in-Tariff (FIT), government capital grants etc. across a lifespan of 25 years.
Some very superficial research into farm land rentals suggests a rate anywhere from 200 euro+ per acre per year.
I've excluded operational costs, assuming that maintenance etc. is insignificant.
An average insolation of 2.75 kWh m^-2 day^-1 is assumed.
An overall panel efficiency of around 13% is assumed.
A purchase price of euro 150 per 100 W panel is assumed. The cost of panel installation is assumed to be the same as the panel itself, i.e. another 150 euro per 100 W panel.
The wholesale spot price for power is something around 0.05 euro per kWh at the moment. This very rough number is assumed to be the base price paid for power.

With these things in mind, the model set-up assumes a small farm of 30 acres. It's worth noting that the actual area is unimportant - the costs etc. all scale linearly with area, i.e. the calculations could be presented for a single 100 W installation and multiplied up. The assumption about connection and infrastructure costs is less valid for smaller installations and the original question was about utility scale installations. Nevertheless, the lessons are more or less valid for all solar installations, whatever the size.

Area: 30 acres
Area: 121410 m^2
Solar AEP: 10561658 kWh
Total number of panels: 80940
Annual rental for use as farmland: 6,000 euro (for 30 acres)

The first run of the model shows the effect of changing the FIT, all else equal. The total price paid for energy increases from Euro 0.05 Euro/kWh, i.e. zero FIT, to 0.19 Euro/kWh, i.e. a FIT of 0.14 Euro/kWh. The results are:


A	B	C	  D	E		     F	        G
0.00	0.05	528083	  46	-11079927    150000	-74
0.01	0.06	633699	  38	-8439513     150000	-56
0.02	0.07	739316	  33	-5799098     150000	-39
0.03	0.08	844933	  29	-3158683     150000	-21
0.04	0.09	950549	  26	-518269	     150000	-3
0.05	0.1	1056166	  23	2122146	     150000	14
0.06	0.11	1161782	  21	4762560	     150000	32
0.07	0.12	1267399	  19	7402975	     150000	49
0.08	0.13	1373016	  18	10043389     150000	67
0.09	0.14	1478632	  16	12683804     150000	85
0.10	0.15	1584249	  15	15324218     150000	102
0.11	0.16	1689865	  14	17964633     150000	120
0.12	0.17	1795482	  14	20605048     150000	137
0.13	0.18	1901098	  13	23245462     150000	155
0.14	0.19	2006715	  12	25885877     150000	173

Where:

A: FIT (Euro/kWh).
B: Total paid for power (Euro/kWh).
C: Annual income from farm (Euro).
D: Years to pay back initial invesment.
E: Net balance after 25 years, i.e. total income - initial outlay (Euro).
F: Total income over 25 years if rented for traditional farming (Euro).
G: Ratio of solar income/traditional farm income over 25 years.

The second run of the model shows the effect of changing the panel cost, all else equal. This could result from the rapidly falling cost of photovoltaics or from a discount for a bulk order. The results are:

A 	B	C	D	E	    F    	G
0.0	0.05	528083	0	13202073    150000	88
0.2	0.05	528083	6	9964473	    150000	66
0.4	0.05	528083	12	6726873	    150000	45
0.6	0.05	528083	18	3489273	    150000	23
0.8	0.05	528083	25	251673	    150000	2
1.0	0.05	528083	31	-2985927    150000	-20
1.2	0.05	528083	37	-6223527    150000	-41
1.4	0.05	528083	43	-9461127    150000	-63
1.6	0.05	528083	49	-12698727   150000	-85
1.8	0.05	528083	55	-15936327   150000	-106

Where:

A: Cost of solar panels (Euro/W)
B: Total paid for power (Euro/kWh).
C: Annual income from farm (Euro).
D: Years to pay back initial invesment.
E: Net balance after 25 years, i.e. total income - initial outlay.
F: Total income over 25 years if rented for traditional farming (Euro).
G: Ratio of solar income/traditional farm income over 25 years.

The final run of the model shows the effect of a moderate FIT of Euro 0.05 per kWh with changing panel prices, all else equal:


A 	B	C	  D	E	    F    	G
0.0	0.1	1056166	  0	26404146    150000	176
0.2	0.1	1056166	  3	23166546    150000	154
0.4	0.1	1056166	  6	19928946    150000	133
0.6	0.1	1056166	  9	16691346    150000	111
0.8	0.1	1056166	  12	13453746    150000	90
1.0	0.1	1056166	  15	10216146    150000	68
1.2	0.1	1056166	  18	6978546	    150000	47
1.4	0.1	1056166	  21	3740946	    150000	25
1.6	0.1	1056166	  25	503346	    150000	3
1.8	0.1	1056166	  28	-2734254    150000	-18

Where:

A: Cost of solar panels (Euro/W)
B: Total paid for power (Euro/kWh).
C: Annual income from farm (Euro).
D: Years to pay back initial invesment.
E: Net balance after 25 years, i.e. total income - initial outlay (Euro).
F: Total income over 25 years if rented for traditional farming (Euro).
G: Ratio of solar income/traditional farm income over 25 years.

Things look marginal in the near future. The economics change rapidly, as you'd expect; if either electricity prices or supports increase or if the cost of panels decreases, then solar PV farms may make financial sense. If the average SEM price for electricity was doubled to 0.10 euro per kWh, because of the market or a new solar FIT, then things might make sense. If the price of solar panels fall towards one euro per watt, then things again look good.

Solar panel costs have been falling by 7% or more per annum. A rate of 7% equates to a halving every ten years. At this rate - and it's faster than this - the price of a solar panel will cost less than one euro in under five years. In ten, it will be only half of today's 1.50 euro per watt. If electricity prices remain stable or increase, then the breakeven point for solar in Ireland, whether utility scale or domestic and without subsidy, is only around the corner.

These figures are very approximate but it seems clear that, in Ireland, distributed solar will be a good bet soon.

How good an investment this would represent depends on how cheap the panels can be made and how much is paid for the power. In the best of the simple scenarios outlined above, the net total income, i.e. after paying for the panels and installation, is around 25 million euro. This is more or less a doubling of the original capital. This is equivalent to a compound return of about 3% on a principal of 25 million euro. From a scan through the literature, this seems to compare well with profits from coal and gas plants.

At present, solar energy makes sense in Ireland if you can use a good portion of the energy directly on-site, i.e. to offset grid purchases.

Although covering countryside in Ireland with panels may seem perverse, if the land otherwise unusable, then it may not be an entirely awful idea in the near future.

Of course, this simple analysis disregards off-balance sheet costs including emissions and climate change effects.

Niall McMahon

Do Solar Farms Work in Ireland?

Links