flexQgrid forecasts congestions and wants to avoid them in the yellow traffic light phase by using a quota. But what is the quota, how does it work and how is it calculated?
A congestions can occur due to too much simultaneous feed-in (e.g. from photovoltaics) or due to load peaks (e.g. from electromobility or heat pumps). We speak of a congestion when the maximum operating limits of grid components are exceeded. In flexQgrid, we use forecasts to anticipate and avoid such congestions. In such cases, we speak of a yellow traffic light phase. Then the planned feed-in or load is reduced over a short period of time so that the congestion does not actually occur. This reduction is guaranteed by the quota. It describes the share of active power that may still be fed in or consumed without causing a congestion.
Who gets the quota?
The quota is communicated to the relevant actors. These can be smart households with an energy management system or units of aggregators. In order to give households the freedom to decide in which form they want to reduce their power, the quota is pronounced on the grid connection point instead of on individual installations. This allows households to optimally adjust their installations according to their wishes. However, not only the installations of a household, but also those of neighbours can optimise each other. After the quota has been communicated, it is possible to place bids on a secondary trading plattform . The bids convey the willingness to deviate from the reduced power. If a market participant can reduce its power even more for a monetary consideration in order to allow another market participant to follow its original schedule, a trade is concluded. After market closure, the trading results are communicated. Then either the original schedule or the schedule adjusted by the market result must be translated by the actors into control commands to the units.
How is the quota calculated?
Different grid clusters are introduced for the calculation. These can be, for example, all installations downstream of a low-voltage feeder, a secondary substation or a medium-voltage feeder. Predicted grid congestions for a secondary substation evoke a quota among the actors connected in the grid behind this substation. A unit or house is also in a cluster for a low-voltage feeder in addition to the secondary substation. Therefore, the system must work without conflict even if one actor has to comply with several quotas.
In order to distinguish between load-side and feed-in congestion, there is both a load quota and a feed-in quota. The load quota for a given cluster in a 15min interval describes the quotient of the available capacity (difference between the capacity of the grid component and the predicted inflexible power in the cluster) and the installed flexible power in the cluster. If the actor is affected by several quotas, the lowest quota value always applies.
This quota was already successfully tested in the previous project "grid-control". However, it consists of conservative assumptions and therefore leads to a low quota - and thus to high curtailment. Battery storage systems, for example, are included with their installed capacity, although they may not have been included in the forecasts. Furthermore, this quota does not take into account the balancing of flexible generation and flexible load, which also leads to conservative assumptions. Since fewer diverse units were connected in the previous project, this aspect was negligible in the framework at that time. In order to fully utilise the grid and the potential of the flexible units, the "flexQgrid" project developed an extended approach for the calculation.
Extended calculation in "flexQgrid"
In order to include the local balancing of flexible generation and flexible load in the calculation, the forecast flexible feed-in is taken into account in the calculation of the load quota and vice versa. Furthermore, it is not divided by the installed capacity, but by the forecast capacity. This leads to less conservative assumptions and thus to a better utilisation of the available grid capacity. Another advantage of this calculation is the display of the grid traffic light phase. The ratio is 1 when we are in the green phase.
However, this extended calculation also brings new challenges. On the one hand, actors must now adhere to the forecast values, since in contrast to the basic calculation, the calculation is not based on the maximum value (corresponding to the installed capacity), but on the planned values. Secondly, the calculation leads to a cascaded quota calculation. In order to be able to comply with several quota requirements from different clusters, already calculated quotas for a low-voltage feeder are taken into account for the calculation of the quota for the secondary substation. The reason for this is that the capacity (and thus a potential congestion) at the secondary substation is reduced if the system capacities in the subordinate low-voltage feeders are already reduced by a quota. If a congestion is predicted at the secondary substation despite observing the quota in the low-voltage feeder, the units must be further derated in order to avoid this congestion as well. This is realised by multiplying the quota of the secondary substation by the quota for the subordinate low-voltage feeders.
Comparison in simulation and field test
Currently, both calculation methods are being compared in a simulation and initial results are being collected for scientific publications. In the field test starting in summer 2021, both methods will also be compared in practice. This will examine whether the assumptions made regarding the advantages and disadvantages of the calculation can also be proven in practice.