Remote sites that pair solar with their diesel generators often have to switch off much of that solar. In a year-long study of one such site, 55 per cent of the solar was lost this way, and every storage design tested would pay for itself within seven years.
By Henning Rath, Chief Executive Officer, EnerVenue
On paper, the solar array at a remote industrial site could supply 70 per cent of the site’s electricity. When an engineering study modelled it hour by hour through a full year, using figures based on a real site, the array supplied 32 per cent, because 55 per cent of what it could have produced had to be switched off. The cause lies in the way a site with no grid connection keeps its own power supply stable. The same cause is at work wherever a mine, an oilfield or a remote community has built solar beside its diesel generators.
Figure 1. An average day at the modelled site, built from all 8,760 hours of the year. Without storage, a generator runs through the middle of the day and much of the solar is switched off. With a 12 MWh battery, midday solar is stored and used after sunset, and the generators run less in the evening.
A generator has to keep running even when the sun is strongest
With no grid to fall back on, the site’s power system has to match supply and demand at every moment, and the diesel generators are what hold the voltage and frequency steady. The solar inverters follow the generators’ lead, so at least one generator must run whenever the site is running, including at noon.
A diesel generator cannot be turned down very far. According to a 2012 report from the IEA’s Photovoltaic Power Systems Programme, running an engine at low load damages it and makes it burn fuel less efficiently, so generators are usually kept above 30 to 50 per cent of their rated output. The same report notes that spare generating capacity may be kept running as well, because a passing cloud can cut the solar output faster than a stopped generator can start. When the array could produce more than the share of demand the generators leave to it, the site’s controller turns the solar inverters down, and that energy is lost. Engineers call this curtailment.
A study of a remote site shows how much solar is lost
EnerVenue’s engineers carried out the study and scaled and anonymised the site’s figures. The site uses 25,000 MWh of electricity a year, around the clock, and diesel still supplies more than two-thirds of it. In the study its ten generators burn about 4.6 million litres of fuel a year, and at US$0.85 a litre delivered, fuel makes up nine-tenths of the cost of running its power system.
A battery puts that solar to use and reduces the number of generators
A battery helps the site in four ways. It stores the midday surplus and releases it after sunset, when the generators would otherwise burn diesel. It can also carry the evening peak, which is what sets the size of the generator fleet, so a new site can be built with between six and eight generators instead of ten. It responds within milliseconds when cloud cuts the solar output, so fewer generators have to be kept running in reserve. The generators that remain run for fewer hours and need fewer overhauls, and at a remote site each visit can mean a long trip for a technician.
The study tested five configurations over 25 years. With the array at its designed size, batteries of 6, 12 and 25 MWh cut diesel use by 13 to 40 per cent. They lowered the levelised cost of electricity, which is the lifetime cost of each kilowatt-hour, by 10 to 22 per cent. Two further configurations in the study also enlarged the array, to 15 and 20 MWp, and took the fuel saving as high as 73 per cent and the cost saving to 29 per cent. Each would pay back its extra investment within seven years, and the smallest battery would do so in two.
Figure 2. The five configurations against the same site without storage. Payback is the time it takes cumulative savings to cover the extra investment, before discounting.
The study costs each configuration as a new build. In the 12 MWh case the site needs seven generators instead of ten, and the three it does not buy cover a large part of the battery’s cost before any fuel is saved. For a remote site, storage belongs in the design of its power system from the start, because it changes how many generators the site has to buy as well as how much fuel it burns.
The study also ran each configuration 10,000 times with 18 inputs, among them fuel and equipment prices, varied at random, and none of the 50,000 runs lost money against the same site without storage. Even with diesel at US$0.65 a litre, the smallest battery would pay back in 2.8 years.
For a remote site, storage belongs in the design of its power system from the start, because it changes how many generators the site has to buy as well as how much fuel it burns.
One generator still runs at all times in every configuration. Handing its job of holding the voltage and frequency to a grid-forming inverter would let the site use more of its solar, so the study’s savings are on the cautious side.
A battery at a remote site has to cope with heat, distance and fire risk
Many remote sites are hot by day, and in that heat a lithium-ion battery usually needs a refrigerated cooling system that runs for as long as the battery does. A fault in its pumps or refrigerant circuit can take the battery offline, and the nearest technician may be several hours away. The battery often sits near fuel storage and processing equipment, where a flammable electrolyte calls for fire protection and separation distances that are hard to provide on a crowded site.
The duty is also demanding, because the battery in the study fills and empties on most days. Every saving in the study assumes that the battery is bought once and still performs to its rating after 25 years of that work. At a remote site, endurance is therefore the first thing to look for in a battery, ahead of its price per kilowatt-hour.
EnerVenue developed its Aqueous Metal Cell for this kind of site. The containerised Energy Prism operates from -20°C to +60°C on forced air, with a small air-conditioning load in the hottest months and no liquid cooling circuit to maintain. The cell uses a water-based electrolyte, carries zero risk of fire from thermal runaway and needs no fire suppression system. It is rated for 30,000 cycles at full depth of discharge over 30 years, at up to three a day, without capacity having to be added. One cycle a day for 25 years uses less than a third of that rating, which leaves room for more daily cycles as the site’s load grows.
At a remote site, endurance is the first thing to look for in a battery, ahead of its price per kilowatt-hour.
A major Chinese oil and gas producer has ordered 26 Energy Prism containers, 11.5 MWh in total, to store solar power at one of its oilfields, and the first units are due to ship in December 2026. In Western Australia, the regional utility Horizon Power is deploying the battery on the microgrid that supplies the Pilbara town of Nullagine.
The result at each site depends on its fuel price and its load
Results will differ from site to site, and the largest factor is the price a site pays for delivered diesel: a higher price than the study’s US$0.85 a litre makes every configuration pay back sooner. A site whose demand peaks at midday uses more of its solar directly and gains less from a battery. The full study, free to download at enervenue.com/use-cases/remote-solar-diesel-storage, sets out all five configurations and how they were modelled. EnerVenue’s engineers can also run the same model on a site’s own load, solar and fuel data to show what storage would save there.
Henning Rath is Chief Executive Officer of EnerVenue, where he is focused on energy storage designed as infrastructure: endurance, safety and lifecycle economics ahead of short-term optimisation.
Discover more from SolarQuarter
Subscribe to get the latest posts sent to your email.
