Sarmad Hanif

Model predictive control scheme for investigating demand side flexibility in Singapore

Recently, power systems have experienced various changes, the most important one being the increase in the share of highly variable renewable energy supply (RES). To counteract the variability of RES, provision of flexibility from the demand side seems to be a viable option. In this paper, the heating, ventilation and air conditioning (HVAC) system, mostly installed in medium to large sized office buildings, is selected to provide demand side flexibility. A model predictive control (MPC) scheme in a receding horizon environment is deployed to provide an economic operation of the building, while respecting comfort constraints of dwellers. Furthermore, robustness is introduced in the MPC scheme to participate in both the energy and reserve market. Simulations are performed to demonstrate the performance of the developed controller under various price signals. In doing so, the controller is also evaluated with respect to its sensitivity towards economical and technical constraints. The National Electricity Market of Singapore (NEMS) is used as a case study and the most important parameters governing the challenges for integrating demand side flexibility in the grid are pointed out.

Cost Optimal Integration of Flexible Buildings in Congested Distribution Grids

Buildings are candidates for providing flexible demand due to their high consumption and inherent thermal inertia. In the future, flexible demand side reserves may also help to relax the expected higher reserve requirements of the grid due to the presence of renewables. However, this flexible demand might be vulnerable to price signals, as the simultaneous increase in consumption by multiple buildings due to low (high) energy (reserves) price periods might cause congestion in distribution grids. In order to integrate congestion-free energy and reserve provision from buildings, this paper presents two benchmark pricing methodologies: (1) distribution locational marginal prices (DLMP), and (2) iterative DLMP (iDLMP). Both methods deploy convex optimization to obtain an optimal solution of the original problem. Using dual decomposition, a settlement scheme, which efficiently distributes the congestion cost among involved participants, is also presented. Case studies are performed on a benchmark distribution system along with the National Energy Market Singapores price framework. The results prove that both methods optimally remove congestion from distribution grids and have potential to be integrated into the theoretical framework of liberalized markets. Furthermore, as a comparison, it is shown that the DLMP-based prices outperforms existing pricing structures of the distribution grid. Hence, using this scheme, the distribution system operator can evaluate existing tariffs and introduce incentives for price responsive demands. However, to support these methods, the high requirement for information sharing in the DLMP method and/or communication technology infrastructure for calculating iDLMPs must exist in the future grid.

Evaluating demand response in the presence of solar PV: Distribution grid perspective

Flexible load operators are particularly interested in monetary transactions of demand response (DR). However, the integration of the DR scheme into the distribution network results in modification of power flows, which has to be managed by the distribution system operator (DSO). Hence, a coordination must be achieved between these two entities to comply with their individual constraints and objectives. With the integration of highly distributed and variable renewable energy, achieving this coordination becomes an even more important task. In this paper, an optimization-based generic model is presented for evaluating DR in the presence of solar photovoltaic (PV) and flexible loads. The integrated optimal pricing methodology is obtained from the developed framework, which takes into account operational conditions of the distribution grid and flexible loads. The economic and operational efficiency of the DR strategy is evaluated in the presence of (1) various pricing structures and (2) available network topologies. Case studies are performed using a validated building model and actual solar irradiation measurements on a benchmark distribution network. For comparison, liberalized market settings of the National Electricity Market of Singapore (NEMS) are adopted in this paper.

Robust reserve capacity provision and peak load reduction from buildings in smart grids

This paper proposes a robust demand-side control algorithm in a smart grid environment for heating, ventilation and air conditioning (HVAC) systems. A robust model predictive control (RMPC) scheme in a receding horizon fashion is deployed, which optimizes electricity cost and capacity market participation of the HVAC system, while satisfying comfort and operational constraints of the building and utility, respectively. Thermal load uncertainties experienced by the HVAC system are included to perform a realistic assessment of the developed controller. The National Electricity Market of Singapore (NEMS) is used as a case study and the developed RMPC scheme is tested for various price signals and scenarios. Numerical simulation results show the effectiveness of the developed framework to be readily adopted by utilities - interested in realizing a grid-friendly and economically efficient demand response (DR) strategy.

Quantifying the effect on the load shifting potential of buildings due to ancillary service provision

This paper analyzes the pessimistic effect on the inherent load shifting potential (LSP) of buildings due to the participation in the reserve market. A generic model-based optimization approach is deployed, which uses a validated dynamic model along with its experienced external and internal disturbances, to quantify the LSP in the presence of various price signals. The theoretical maximum LSP is obtained using a base energy price without the provision of ancillary services (AS). The deviation from the base case LSP is observed after including the time varying energy and the reserve price from the spot market. Factors affecting the LSP are found to be as: (1) physics of the model, (2) the nature of price signal and (3) the competency of the reserve price with respect to the energy price. Due to its simple formulation, low computation requirements, and modular nature, the method proposed in this paper can easily be deployed by retailers and system operators for the assessment of monetary incentives as well as qualified load type for various demand response (DR) services in liberalized markets.

Combination of renewable generation and flexible load aggregation for ancillary services provision

Liberalized markets provide opportunities for load aggregators to reduce their operating cost by participation in the wholesale electricity market. Reduction in operating cost result from shifting flexible loads based on the electricity price and also from incentives paid for ancillary service provision. Increased penetration of distributed renewable generation could help distribution system operators (DSO) reduce energy losses and increase network reliability. On the other hand, variability of renewable sources introduce additional challenges that need to be addressed in order to reduce their adverse effects on the power system. This work proposes an energy management system for aggregation of controllable loads in the distribution system. Local generation uncertainties are considered using model predictive control (MPC). The goal of the controller is to provide a schedule for the controllable loads while considering both local solar photovoltaic (PV) generation and provision of ancillary services. A case study based on the National Electricity Market of Singapore (NEMS) is presented to validate the proposed method.

Cost-optimal operation for a flexible building with local PV in a Singaporean environment

In the future, flexible demand with the combination of renewable energy may hold the key for a sustainable power grid. However, if not managed properly, this combination may pose a threat to the reliability of the grid. In this paper, we analyze the effect on the cost-optimal operation of the flexible building in the presence of locally connected solar photovoltaic (PV) system. Based on a detailed thermal model of a building, an optimization based operational strategy has been presented. The objective of this strategy is to optimally schedule energy and reserve in the presence of flexible loads and the PV system. Two forecasting methods, in a rolling horizon fashion, are deployed to evaluate the interaction between the uncertain PV output and the operation of the whole system. Case studies are performed using the pricing framework of the National Energy Market Singapore (NEMS). Results show that a trade-off exist between the cost optimality of the overall system and the excess PV output fed to the grid (reverse power flow).