11:10 - 13:00
Submission 159
Ensuring reliability of FCR and aFRR provided by Wind & Solar
WISO25-159
Presented by: Julien Callec
Julien CallecClarisse DelauneSamuel RatouitFrederic Paugam
RTE, France
The transmission system operator (TSO) must guarantee network stability, today and considering future developments. For example, the TSO must ensure that it always has the necessary reserves to maintain frequency within acceptable ranges, as well as the assets capable of delivering these reserves with the correct dynamics. As wind and solar generation increases in the energy mix, renewable energy producers must increasingly contribute to system flexibility. In France, for example, some of them are currently required to have the capability of providing automatic frequency control (frequency containment reserve (FCR) and automatic frequency restoration reserve (aFRR)).

For providing FCR and/or aFRR, wind and photovoltaic farms will operate below their optimum operating point, Pavailable(t). If they participate in the delivery of a symmetrical frequency service, they will have to reduce their power output to provide FCR and/or aFRR around a base power Pbase(t), where Pbase(t) = Pavailable(t) – Scheduled_ReserveFCR – Scheduled_ReserveaFRR with Scheduled_ReserveFCR and Scheduled_ReserveaFRR the amounts of scheduled reserves. If the regulation is only downwards (as it is allowed for aFRR), Pbase(t) can be equal to Pavailable(t). However, in both cases, symmetrical and unsymmetrical regulations, it is necessary to know the value of the optimum operating point, Pavailable(t), at each time in order to compute a reliable setpoint Psetpoint(t) that includes the participation of the renewable asset in frequency control PFCR(t) and PaFRR(t), as Psetpoint(t) = Pbase(t) + PFCR(t) + PaFRR(t). Once the set point has been established, the power must then be released by the wind or photovoltaic farm with the correct dynamics.

The main issue here is that, when the wind or photovoltaic farm is operated below its optimum operating point, the value of Pavailable(t) cannot be measured and must be estimated. As with any estimate, there is a margin of error on the estimated value and this error will be reflected in the setpoint calculation Psetpoint(t). Inaccurate computation of the setpoint could result in insufficient delivery or availability of the reserve on the asset. As an example, if we consider a wind farm providing only symmetrical aFRR, the effective available reserve upwards Real_ReserveaFRR will be equal to Preal(t) - Pbase(t), where Preal(t) is the effective power that can be extracted at time t from the wind, but if Preal(t) << Pavailable(t) then Real_ReserveaFRR << Scheduled_ReserveaFRR, the amount of scheduled aFRR.

This paper aims to address this issue based on an internal statistical study of lab data. First, it assesses the impact on French reserves if no action is taken to address the estimation error in Pavailable(t). Then, it presents the solutions that RTE is considering to secure its reserves, taking advantage of renewable energies' arrival as new key players in power system stability.