All agents participating in the electricity market commit to generate or consume an accurate amount of energy in a precise future time interval (imbalance settlement period). When the final electrical energy provided (generators) or withdrawn (consumers) from the system differs from the one initially programmed and agreed upon with the system, what is known in the sector as a deviation of electrical production occurs. Both the generator, that produced more (or less) energy than promised, and the consumer, who ultimately needed more (or less) energy than it said it would, will be economically penalized due to the extra work that the operator has to carry out to adjust the system and ensure a correct balance between generation and demand, as well as ensuring a correct supply of electricity.
The imbalance management is key to the profitability of any agent in the electricity market. There have been repeated conversations about imbalance from the generators point of view. Nonetheless, in this article we will focus on deviations from the point of view of the consumer and its impact on the electricity bill. We will talk about what they are, what impact they have and what it can be done to mitigate their economic cost overrun, all explained through a practical example to facilitate the reader’s understanding.
What is exactly and electrical imbalance?
The imbalances are defined by the electrical system operator (REE in Spain) as the difference between the energy measured in busbars and the energy programmed in the market.
DESVu = (MBCu – PHLu)
For consumers, the term MBCu refers to the consumption measured at the end-point (downstream of the meter) raised to power station busbars. The term PHLu refers to the energy hourly program settled with the market operator (OMIE in Spain and Portugal). In the electrical system the electrical losses associated to a consumption end-point (losses upstream of the meter) must be covered by itself. Since the final losses percentage is not known until long after the physical delivery of energy has taken place, to calculate the PHLu it is necessary to estimate in advance the losses that each end-point will have associated (REE periodically publishes some loss forecasts based on tariffs and periods). As a general idea, points of consumption with high voltage (tariffs 6.X) can have associated losses between 2 and 7% of their downstream consumption while for low voltage points (tariffs 2.X) these numbers can double up to 14%.
The system operator always refers to imbalances as from the point of view of the producer or generator. Two types of deviations are defined, the “EXCESS” imbalance and the “DEFICIT” imbalance (See Table). It is important to understand the sign of deviations because, although we will focus on the deviations from the point of view of the consumer, the nomenclature remains constant.
It is convenient clarifying that the collection rights or payment obligations associated with the imbalances will be settled with the system operator (REE) and not with the market operator (OMIE). REE since 2006 manages its settlements through MEFF (the governing company of the Spanish financial options and futures market), which will act as the central counterparty of the collections and payments. In addition, when generating payment obligations due to imbalances it is understandable that economic guarantees are established (as is the case with the market operator) and which must be deposited by the market participants in accordance with the provisions of REE. More details on operator guarantees requirements here (in Spanish).
Imbalance costs and the electricity bill.
The costs of incurring in imbalances is calculated through the official procedure defined in the Spanish Official Bulletin (BOE) available in this link (in Spanish). The cost of a settlement subject (market subject, SM) due to imbalance depends on its magnitude, its sign, the system status and the Cost of excess production imbalances (CDB) and Cost of production deficit imbalances (CDS), published and updated by REE. These costs are calculated by REE after the time imbalance settlement period. Their value will be higher or lower depending on the measures that have been necessary to activate by the operator in that time interval, to keep the system stable and supply energy to all participants.
The measures to which we refer are included in the imbalance management procedures, a system adjustment service provided by all programming units previously authorized for the provision of this service. The higher the sum of the imbalances from the participants in the system, the higher the adjustments that must be made and the higher the imbalances cost. This implies that the entire cost of the imbalance management derived from bad planning of agents or undesired unforeseen events, is shared among all market participants. For example, if a nuclear power plant suffers a breakdown and has an unplanned stop, the cost of correcting such deviations will be paid by all participants in the market. If in a region there were predicted wind speeds suitable for generation, but finally there was nothing but calm, the cost of adjusting the system will be paid by all.
Once the imbalances costs for a specific market (SM) subject has been calculated, it is then allocated among the programming units it comprises in accordance with the aforementioned official documentation (PO.14.4). Thus, the programming units that have contributed to mitigate the total imbalance of the SM will not be penalized. Not being penalized in this context means that the units will pay or be paid for the energy imbalance at the same price they would have obtained had they previously nominated correctly. However, the units whose imbalances have contributed to increase the total imbalance of the SM, will suffer a penalty. This penalty will deviate the price that they pay (increase) or receive (decrease) for the sale of energy.
How can consumers reduce their imbalances costs?
Once it has been presented what the imbalances costs are, why they exist and how much is its cost, we analyze some of the measures that can be adopted by consumers in order to reduce them and achieve savings in the energy bill.
Aggregate units to increase the portfolio effect.
Add units in the same subject to increase the screening effect between them. By having several units, the “excess” imbalances will be mitigated with the “deficit” imbalances and vice versa. The more units the larger the shielding effect will be. For example, for a retailer, who manages the purchase of thousands of customers, the portfolio effect is strong. In the end, the units or clients that have behaved well with respect to their detours will help those whose forecasts led to higher nomination errors.
Improve energy consumption forecasts
Knowing how much energy will be consumed in the future is not easy, even if the time horizon is as short as 24h. The weather, the day of the week and the production scheduled are, among others, are factors that directly influence the energy demand. To better understand the relationship of these parameters with the final energy consumption efforts must be addressed to build accurate prediction models. These models will help improving the purchase and schedule of everyday energy. With a more adjusted program the imbalance is smaller and the economic savings larger.
If there is not space to invest in mathematical or machine learnings models that help the consumer improve their forecasts, there should be at least an active management of the energy consumption programs. For example, reacting quickly to a technical problem that might stop a plant or lower down its activity can lead to significant savings if the purchase if modified on time. Furthermore, with Intraday markets and the new European continuous market, making changes is now more accessible than ever.
Come forward your losses
As mentioned in the beginning of the article, the consumption that is considered in the calculation of the imbalances costs is the consumption raised to power central busbars. To calculate this value, basically two data must be known: one, the consumption at the end-point, and two, the losses associated with that end-point, meaning the energy lost on its way from the generator to the customer. Losses are published by REE but these are usually rough estimates at the beginning. From the time of publication to their final value they can vary substantially. Being able to use loss coefficients closer to reality in the nominations will lead to a more controlled imbalance. If we systematically have errors with the losses used to raise our consumption forecasts when buying our energy, we are increasing our costs.
Install storage systems
In the wind and photovoltaic farms, it is common to install storage systems to neutralize the imbalances. However, at the consumer level it is more complicated because the energy volumes are much lower. With the progressive reduction of storage systems costs, it is not impossible that large consumers and retailers may consider the possibility of using batteries in the future to reduce the imbalances costs.
A case study is presented to help understand the imbalance management and its impact on the electric bill. Let’s suppose we are a direct consumer participating in the electric system through a market subject with the name of GOT. Suppose also we group under this SM two programming units, BRAAVOS and MEEREEN, both energy consumers. We will analyze our imbalances costs for a specific month and we will calculate what effect the correcting measures will have.
The following table shows what our consumption and energy program have been along this month. It shows that, on average, we have over forecasted our program, therefore, incurred electricity production imbalances.
As some examples of imbalances costs calculations, the following table shows a sample of three hours for the analyzed month for our subject GOT. In the table CDS and CDB are shown together with the state of the electrical system, meaning long for more generation than demand or short for more demand than generation. Depending on the sign of the SM imbalance respect to the state of the system there will be a penalty.
At hours 11 and 13 the system is short, that is, more production and less consumption are needed for balancing the system. Therefore, it will be penalized if a consumer demands more energy than programmed. Our market subject, at the 11th hour, had an Excess Imbalance, aligning with the system because it consumed less energy. For this hour, we will be paid by REE for the energy not consumed at the market price of that hour, that is, €61.48 / MWh. A collection right of €11.89 was generated (our imbalance was 0.19 MWh). We have behaved well, and we have not been penalized, for us this imbalance has had no cost because it is the same payment we would have done if in the nomination of the previous day we would have nominated perfectly, buying 0.19 MWh less for this hour. However, at the 13th hour the story is different.
The system is still short but for this hour we have consumed more than what we have committed to, incurring in a deficit imbalance, which goes against the system. REE will penalize us. The cost published for deficit imbalance at this time (CDB) is 10.88 € / MWh. Because of these 0.33 MWh that we are demanding extra, we must pay this energy at a rate of 66.89 €/MWh (OMIE price + CDB), generating a payment obligation of €22.07. If the previous day we had nominated perfectly, that is, buying 0.33 MWh more than we bought, we would have paid for this energy 0.33MWh * 56.01 € / MWh = 18.45 €. The difference between what we would have purchased at an earlier time, those € 18.45, and what we must pay now, the € 22.07, is the cost associated with the imbalance at hour 13 of this day.
The sum of all the hourly costs in the month indicates the total imbalances cost for that month. It should be noted that if for a certain hour our cost is 0 (see hour 8 in the example) it does not imply that we have not generated a collection right or a payment obligation. What happens in this hour is that we have deviated from our program 0.21MWh in favor of the system and, although there is no penalty, we obviously must pay for that energy. In this case there is no difference money wise between having previously nominated perfectly having an imbalance, both scenarios we end up paying pay 0.21MWh * 60.84 €/MWh = 12.78€. This should not be considered as an imbalance cost.
Since we group two programming units, we will analyze the portfolio effect due to having participated in the market under the same subject. In the following graphs, firstly, the imbalances costs that the units would have had having attended the market individually. In the second, the costs have been allocated to both programming units according to the process commented in the PO.14.4 of the abovementioned official Spanish document. They show how the sum of the two units has been reduced considerably thanks to the screening effect, going from paying in imbalance costs €2930 to the €2163 of SM_GOT (equal for both graphs).
The most common metric used when referring to imbalances costs is the rate €/MWh, calculated dividing the total imbalances cost by the MWh programmed for a specific month. It gives us a clear idea of how well we are doing our imbalances management. In the previous case of portfolio effect costs, let’s see what these rates are:
Our efforts in the imbalance management strategy should be aimed at reducing this rate. It is hard to give specific values for this parameter to know if we are doing it right or wrong because it depends on the type of consumer, the industry to which it is dedicated, the month, etc. In general, values like the one we have seen for GOT can be considered normal. As consumers and generally speaking, values between 0.15-0.30 € / MWh can be considered good, values between 0.3-0-70 are acceptable, but show that there is probably room for improvement if we perform an active management of deviations, and values that exceed 1 €/MWh should make our alarms jump and review our nomination processes. We must in any case make a thorough monitoring of this parameter and react to any sudden change from one settlement to another.
We are now going to analyze if we could have refined our consumption predictions more accurately to reduce imbalances throughout the month. We simulate that our hourly predictions for each program unit have been, randomly, between 5% and 15% more accurate and, therefore, our imbalances have decreased. This improvement in predictions translates into a direct saving in the imbalances costs. If we had used these predictions, we would have spent €1929 per imbalances costs, meaning a reduction of 10.8%. In addition, the fact of using more precise estimated losses to raise our consumption to power central busbars will have a similar effect because it will improve the program to nominate and reduce the net imbalance. All these measures are accumulative and, applied together, will probably lead to significant savings in our energy procurement costs.
We hope with this article to have offered an overview of the electric imbalances costs and their importance from the point of view of the consumer. It should be clear by now that an active imbalances costs management strategy is crucial, and every consumer agent should dedicate resource to optimize it to reduce costs. Specially these days with the OMIE spot price exceeding €70 on daily averages.
Any improving in the imbalance costs ratio €/MWh, whatever its magnitude, is more than welcome.
Álvaro Picatoste Ruilope | Analyst