09-10-2020

Michel Forget: accurate baseline data about the project is key to quality solar forecasting

The GetMarket platform interviewed Michel Forget, Commercial Director at IKNET. We talked about the current problems of Ukrainian solar energy, high-quality forecasting of solar power generation and the future responsibility of green energy producers for imbalances.

- As you know, the recently adopted Law 810- IX has quite dramatically increased the level of responsibility of RES producers for the generation schedule of their power plants. What do you think, is the industry ready for this?

- Indeed, starting from the next year, all RES stations (members of the balancing group of the Guaranteed Buyer) will be responsible for deviation of hourly output forecasts from the real hourly generation. For solar power plants over 1 MW it will be 50% of the cost of unbalanced energy in case of a deviation of more than 5%.

Of course, in order to "guess" the forecasted generation within a 5% error, it is necessary to use specialised software products. Nowadays, almost no one uses such forecasting products in Ukraine, and not even to mention the fact that not all solar power plants are equipped with their own weather station. Thus, today the industry is not ready for this. The positive point is that there are 3 more months to prepare.

- What is the reason for this situation? Would it be worthwhile for RES producers to prepare in advance for the fact that sooner or later they will need quality forecasting?

- The reason is simple - the question of penalties for imbalances has only appeared in Ukraine now, which is 10 years after the start of green energy development. Therefore, it is quite natural that the forecasting of solar and wind power generation is poorly developed in Ukraine at present. However, the situation is changing rapidly now and, as I have mentioned above, the RES market players have another 3 months to prepare. The positive thing about this situation is that over the past 10 years very advanced products for forecasting have appeared and developed in Europe, so "reinventing the wheel" is not necessary!

- It is clear that in such a time period, companies need to make quick decisions about the choice of software. What should companies know about products that are on the market?

- The best thing is, of course, to compare several products, both foreign and Ukrainian, in trial mode. However, there is really no time left for this today. My recommendation is to trust the world leaders, who service hundreds of projects around the world and have been continuously improving their forecasting accuracy over the last decade. At the same time, I am confident that in a few years from now Ukraine will have its own products, which will be able to compete.

- IKNET is already working with companies in the area of solar plant generation forecasting. In your experience, what should be paid special attention to?

- Yes, our company IKNET is the official representative of Slovak company SOLARGIS in Ukraine. SOLARGIS is one of the world leaders that I mentioned before.

Over the last few months, we have been running about 20 projects in different regions of Ukraine in trial mode. The main conclusion we have come up with over this time is the key importance of providing accurate baseline information about the project. To give you an example - during the first week for one of our projects we noticed a constant and stable deviation of real hourly generation from the forecast at around 40%; although Solargis products are very accurate and mainly give an error of no more than 5%. As a result, after double-checking, it was found out that the customer had mistaken the north and the south when filling out the questionnaire! Of course, this is a somewhat ridiculous case, but often there is a deviation of a couple of degrees in the real orientation of the panels from the design; sometimes the angle of the panels or the distance between rows and between the panels' tables is inaccurate - all of these affect the calculation of the initial configuration of the power plant and the calculation of losses from shading. With rooftop power plants, it is even more difficult - in a single project, the roofs can be not only gabled (and this is two different azimuths), but also be of different sizes, with different angles. We have already experienced a project where the forecast model is based on five different configurations.

For accurate forecasting, it is absolutely essential to provide accurate technical information about the project, as well as regular data from the meters and the local weather station.

- What guarantees should be expected from forecast software suppliers? After all, a producer can incur financial losses if a forecast is not accurate.

As far as I know, none of the leaders on the forecasting market give financial guarantees for their products. Which is partly logical - after all, it is simply impossible to calculate now the exact amount of the penalty for unbalance for a particular object without knowing the amount of deviations of all market participants, as well as, the prices of day-ahead market and the balancing market. Therefore, again, I can recommend contacting companies specialising in forecast products that can provide the most accurate forecast. In fact, the experience and reputation of such a company will be a guarantee in this case.

On the other hand, I do not eliminate the possibility that specialised insurance products may appear on the market after a certain period of time, when the mechanism for calculating penalties for imbalances is clearer and better practised.

Before
09-10-2020

Мішель Форже: точна вихідна інформація про проект є ключовою для якісного прогнозування генерації СЕС