17 09 2017

Solar energy is here – and it works!

The cost of solar energy has plunged in recent years, and solar panels are now dirt cheap. We have just installed a 46 kW solar roof top system here in Leon, Nicaragua, which produces energy at 7.5 cents of a US dollar – that is around a third of what the electricity company charges us. So there is no excuse for continuing contributing to global warming. The energy revolution is here, right now.

Solar energy seems to be unstoppable, driven by the steadily falling cost of solar panels. I posed the question two years ago, whether this was the turning point for fossil energy. We can now say that, yes, it was the turning point. When I installed my first small solar system 5 years ago, the solar panels cost me USD 1.40 per watt. The solar panels for our new system, that we have just installed, cost me USD 0.48 per watt. And costs are still declining.

Some years back, the solar panels were the expensive part of a photovoltaic system. Not any more. Unfortunately, the solar charge controllers and the inverters that you need to convert the direct current from the panels into 120/240V alternate current have not dropped in price, at least for small commercial systems as ours, so they are now the expensive part of the system. The total investment cost for our system is USD 1.34 per installed watt solar capacity, distributed as shown below (we have imported the solar panels and accessories directly from the factory in China, while we bought the inverters and solar charge controllers in the US):

It is quite sunny here in Nicaragua, so the panels produce a lot of energy. Data from the last three years from our existing small 4.2 kW system show that on average the daily production is 4.27 kWh per kW installed capacity. That is quite good: more ore less the same as in Southern Italy or Arizona. Of course, production varies during the year. The best months are January to March (the dry season), and the months with the lowest production are from June to September (the rainy season). But the monthly average never drops more than 15% below the annual average, so it is fairly stable over the year.

The operational costs of the solar system are negligible. In the dry and dusty months from January to May, we have to clean the panels once or twice a month. That is done with a hose and a soft brush. The rest of the time it is just humming along by itself.

The inverters (left) and the solar charge controllers (right)

The cost per kWh produced depends therefore almost entirely on the capital costs, that is, on how long the different parts of the system will last, and then of course the interest rate. The solar panels should last 25 years – actually they should be able to continue producing after that, but below 80% of the original capacity. The inverters and solar charge controllers have a shorter life span, somewhere between 10 and 15 years. If we use an interest rate of 5% and assume the equipment to last 15 years before it has to be renewed, the cost per produced kWh is as mentioned 7.5 cents of a USD (taking into account the gradual decline in the production of the solar panels over the years). The price we are paying presently to the electricity company is 24 cents of a dollar (flat rate, no smart metering). So apart from contributing to combat climate change, which is our main objective, it is actually also a good investment.

Large scale commercial solar farms can produce cheaper than we can. I guess the main reason is economies of scale for the charge controllers and inverters. According to a recent report from the US Solar Energy Industries Association’s, the cost of large scale solar farms in the US is now around 1 USD per watt installed capacity, that is, around 25% lower than our system. In China and India the cost is even lower than that. The puzzle is still, why smaller PV systems, as ours, are so expensive in the US compared to e.g. Germany (where the cost is only a little higher than ours). It seems to come down to inefficient installers in the US, charging a too high margin.

34 of the solar panels are phasing East to capture the morning sun.


Well, back to our own system. It is well-known that even if solar energy is dirt cheap, it is not stable. You only have energy during day-time and production depends on how cloudy it is. Most solar systems therefore deliver the surplus production to the grid and buy back from the grid during the night or when production is insufficient. What the electricity company pays for your energy (the feed-in tariff) is normally much less than what it charges for providing you with energy. Fair enough, as they have to secure back-up capacity and pay for the cost of the distribution network. However, here in Nicaragua the electricity company does not buy your energy at all – according to the local newspapers the meters even count the reverse current, so you will be charged for what you deliver too. We have therefore designed our system basically as an off-grid system with batteries, and we then use the grid for back-up when we have neither solar, nor energy in the batteries. A new law has just been approved that requires the electricity companies to put up bi-directional meters and pay for the electricity delivered to them, but the regulations yet to be published.

Mounting the panels on the main building

Putting panel number 164 - the last - into place.
The inverters are heavy....

For those who have their main electricity consumption during day-time, a solar system is, as explained above, an obvious investment. As our system is for a small hotel, most of the electricity consumption is unfortunately during night hours – around 35-40% is used during day-time and 60-65% during the evening and night (mainly air-conditioners). To be able to cover a substantial part of our electricity consumption from solar, we therefore need to have capacity to store electricity so we can harvest solar energy during the day and use it during the evening and night.

Energy storage is still a headache. You can get very good lithium-ion batteries that will do the job, but they are expensive and most will only last 5-7 years if you charge and discharge them on a daily basis. You can get longer lasting lithium-ion batteries, but they are still very expensive. We have looked through all the options for different lithium-ion technologies, even lithium-titanate, and made the calculations, but none would pay off. The same is the case for the emerging flow-battery technology. We have therefore finally chosen a very old-fashioned technology: Nickel-Iron batteries, imported directly from the factory in China. The technology dates back to the second half of the nineteenth century, where it was patented by Thomas Edison (the inventor of the electric bulb) and used among others in electric cars. It has some draw-backs: it has a high internal resistance and is therefore rather inefficient, and it needs a lot of watering. But it has one big advantage: it should be able to last up to 30 years, even if that means changing the electrolyte now and then. And it is known to be very robust to over-charging and -discharging. Furthermore, there are no polluting elements in the batteries. The recyclers will happily take the nickel and the iron, and the electrolyte is potassium hydroxide (KOH, “caustic potash”, traditionally used for making soap), which can be neutralized with acid and discharged. The cost of our batteries was 337 USD/kWh delivered (CIF) in Corinto, Nicaragua’s main port, which is more or less the same as standard lithium batteries (LFP), but without battery management system (BMS).

We have built a nice house for the batteries in the garden.

And how is our system performing, then? Well, it has worked for one and a half month now, and does the job as expected. We have a challenge with the battery bank, as it needs an initial charging that we have not been able to do with our Schneider charge controllers, as they have a software limiting the charging voltage to 64V (despite claiming 67V in their specification sheet). This limitation is put in to protect the commonly used lead-acid batteries, which will suffer damage if overcharged. I am still discussing this with the Schneider engineers and I am crossing my fingers that they will provide me with a software update so I can increase the charging voltage. They are very cooperative, but not particularly fast, alas. If they don’t solve it, we will find another solution. The battery bank presently delivers around 60-70% of the expected capacity only because of this.

Even so, we just got the electricity bill for August, the first full month of functioning of our new system, and the saving was more than 900 USD. And that is for a month with relatively little sunshine and relatively low occupation at the hotel. That is enough to justify our investment. So far, so good. For the environment – and for us.


Some data on our system: We have 164 Yingli monocrystalline 280W solar panels, 3 Schneider XW+ 6848 120/240V inverters and 10 Schneider MPPT80-600 charge controllers, plus some additional equipment for remote monitoring and communication. Our battery bank consists of 120 1,000 amp Ciyi Nickel-Iron batteries for a total capacity of 144 kWh. For those interested in the performance of the battery bank, I shall come back with an update later on.


Read 241 times
Rate this item
(0 votes)
Thorbjorn Waagstein

Thorbjørn Waagstein, Economist, PhD, since 1999 working as international Development Consultant in Latin America, Africa and Asia.

Related items

More in this category:

Leave a comment

Make sure you enter all the required information, indicated by an asterisk (*). HTML code is not allowed.