Chris Caerts, Product Manager at VITO, detail the FHP (Flexible Heat and Power) dynamic coalitions of distribution grid connected power to heat resources project and how this could provide local and system level services throughout Europe
The power system is undergoing a fundamental transformation, driven by the ever-increasing share of renewable energy sources (RES), like wind and solar. These renewable energy sources, that are replacing traditional fuel-based and nuclear generation, are integrated both at the central system level (e.g. large off-shore wind farms) and at local distribution grid level (e.g. building level PV). As the amount of such intermittent renewable generation increases, so does the risk and amount of RES curtailment.
The FHP project has developed a standard-based multi-agent software framework that allows the mitigation of RES curtailment by shifting – in a coordinated manner – the electric consumption of distribution grid connected power-to-heat conversions, taking the local grid conditions into account.
Two types of RES curtailment can be distinguished. Technical RES curtailment is typically caused by distribution grid-connected PV generation: as more local PV is added, this may lead to reverse power flows for which the grid was not designed, causing congestions and voltage problems. Economic RES curtailment results from market conditions when there is (forecasted) excess generation, which is not in balance with the (forecasted) consumption. There is a clear need for the cost-effective mitigation of RES curtailment: which avoids wasting emission-free energy, increases the amount of (OPEX) free energy in the energy mix – which potentially reduces energy prices – and has a positive effect on RES investment business cases.
In the FHP project, the distribution grid connected thermal storage and flexibility (provided by thermal inertia) is used to shift heat pump power-to-heat conversions in buildings and large (seasonal) heat storage vessels in a RES curtailment mitigating manner. A dynamic coalition manager was developed that coordinates the operation and control of a local cluster of heat pumps, to ensure that flexibility activations do not cause any local grid problems. For this, a USEF (1) inspired transaction scheme was developed and implemented as a multi-stakeholder multi-agent platform that facilitates optimal flex trading between flexibility sources (e.g. clusters of buildings represented by a dynamic coalition manager), market operators (e.g. BRPs) and the local grid operator (DSO). RES curtailment mitigation business use cases have been implemented that operate at the day-ahead timescale, the intra-day timescale, as well as real-time (acting on intra-ISP imbalance forecasts).
To characterise the thermal flexibility of buildings, dynamic thermal grey-box models are created in a replicable human expert-free manner, solely based on measurements. Machine-learning techniques are deployed for multi-zone characterisation without any architectural information (a zone is associated with each temperature sensor). Current experiments indicate that with 10 days of auto-validation followed by 10 days of auto-correlation, sufficient accuracy for single-family residential buildings can be obtained. This will be further validated through a pilot validation in Sweden (winter 2018). These building dynamic thermal models are complemented with measurement-driven black-box heat pump models.
Using these two models, a building level power consumption and control profile is calculated for each of the participating buildings. This calculation can be a (thermostatic) on/off control or PI control, but it may as well be the result of optimisation that minimises energy consumption (e.g. minimising losses by smarter pre-heating) or energy cost for a specific business case (e.g. dynamic tariffs, peak pricing, solar self-consumption etc.). This (optimal) consumption plan of the flexible controllable consumption is added to the forecasted non-controllable consumption, to create a baseline consumption plan for each building.
Next to this, the remaining/updated flexibility with respect to the calculated (optimal) baseline is characterised as a Flex Graph that describes the upper and lower power consumption profiles – that do not violate comfort settings – in the function of time. Both the baseline consumption plan and Flex Graphs are communicated to the dynamic coalition manager who aggregates this information for all participating buildings and sends it to the DSO. The DSO uses the aggregated baseline information to check whether it would cause grid violations and if it does, determines an optimal flex activation plan within the provided Flex Graph.
In the FHP project, the currently used prime objective is the minimisation of local RES curtailment. This will be extended in the future to minimise cost, taking into account curtailment (compensation) costs, flex activation costs and optionally grid losses.
This optimal flex activation plan is sent back to the dynamic coalition manager who disaggregates it into flex activation requests for each of the contributing buildings. This disaggregation uses an iterative incentive based negotiation that results in an aggregated response of the participating buildings – that is as close as possible to the requested flex activation plan.
The DCM also acts as a local energy community manager, that uses the available aggregated flexibility to offer RES curtailment mitigation services to balance responsible parties (or system level RES owners). If such parties are confronted with market conditions that would lead to a RES curtailment decision, they can instead bilaterally engage with the DCM to ask for a consumption shift/increase as an alternative. Such a flex activation request coming from the BRP will be accommodated to the maximum possible extent, within the DCM’s aggregated flexibility and results in a proposed aggregated baseline update for the cluster of buildings. This proposed updated baseline, along with the updated Flex Graph, is sent to the DSO for clearing.
The DSO will check whether the proposed baseline would cause grid constraint violations and if so, calculate an allowed baseline within the Flex Graph that is feasible and deviates as little as possible from the proposed baseline. Based on this, the DCM formulates a flex offer to the BRP (that maybe only partially fulfils its request) and if accepted, disaggregates the resulting flex activation plan over the buildings in a similar manner as for the local RES curtailment case.
USEF: Universal Smart Energy Framework (https://www.usef.energy/)
FHP project is funded by European Union under the grant agreement no.731231.
Please note: this is a commercial profile
Chris Caerts
Product Manager
VITO
Tel: +32 14 335912
