(566c) Optimizing Generator Efficiency Using An Optimized Network of Loads

The existing electricity industry is unidirectional in nature. It converts only one-third of fuel energy into electricity and almost 8% of its output is lost along its transmission lines, while 20% of its generation capacity exists to meet peak demand only (i.e. in use only 5% of the time) [1].

Every electrical utility uses different generating resources to meet the total demand. As competition in the utility industry increases, each utility attempts to maximize the value of their resources at different levels of operation. The hierarchical approach for maximizing the benefits of resources is divided into several operation planning models with different time spans and computational burdens.

• Strategic long-term operation planning which covers 1 to 4 years with a monthly time step
• Medium-term optimization over a daily or weekly time steps for up to a year
• Short-term optimization with hourly or sub-hourly time steps

Permanent balance between generation and demand in real time is the key to stable and reliable operation of the electric power system. Electric demand follows seasonal and daily cycles which are predictable using demand forecasting methods. But the very short term and real-time component of system demand has stochastic characteristics imposing mismatches between supply and demand. In order to keep these mismatches within permitted levels while reducing balancing costs, a hierarchy of different balancing tasks is designed in the Energy Management Systems (EMS) of operation centres. Generation resources are committed and dispatched economically to meet demand. Since demand is a stochastic fluctuating quantity, flexible generation units have traditionally been used as balancing tools to follow the demand. As renewable energy penetration increases, the variable and uncertain characteristics of demand become even more severe. This intermittency increases the need for installing new spinning and regulation reserves and sub-hourly load following capacity. Each generation unit has an efficiency characteristic relating the input mechanical power to the output electrical power supplied to the grid. This efficiency characteristic is the most important factor used in generation scheduling and dispatch to maximize the total efficiency of the utilized resources. Meeting the reserve constraints and following the fluctuating demand trends lead to loading of the units in sub optimal operating levels.  

The emergence of new metering technologies and advances in communications and control systems has made DSM a feasible and economic tool to operate smart power grids. The Smart Grid is expected to overcome the serious existing challenges such as the low efficiency of generation and a high carbon footprint [2]. Maximizing asset utilization and high penetration of renewable energy resources and Energy Storage Systems (ESS) are also targeted in the smart grid paradigm [3].

Demand Response (DR) refers to the responsiveness of power consumers to price or control signals from utility. It can be exercised in contingency events as a reliability resource or during price spikes as efficiency resource. The Federal Energy Regulatory Commission (FERC) order 719 [4] and 745 [5] promote participation in DR programs and discuss compensation rules and initiatives. Current DR programs in electricity markets are based on voluntarily or directly controlled load curtailments or load reductions to support the power grid while providing a revenue stream for participating customers and aggregators [3]. DR provides excellent opportunities to avoid/defer  investments required to meet the current fluctuating demand and reduce the overall energy price. However, load curtailments or load reductions cause temporary discomfort for the participants. This inconvenience limits their deployment frequency and duration.

Energy Storage Systems (ESS) such as batteries, super capacitors, flywheels and superconducting magnetic energy storage (SMES) systems etc. have been suggested for facilitating intermittent generation integration, improving grid reliability and power quality, enabling energy arbitrage, transmission distribution investment deference and providing ancillary services in [6], [7], [8] and [9]. The main drawback of using these storage systems with a power electronics interface is their high capital cost, low energy density, low efficiency and potential adverse environmental effects [8]

Inherent storage and flexibility of some residential, commercial and industrial loads make them valuable resources for power system operators dealing with active power manipulation. The capability to aggregate and continuously control individual responsive loads based on power system requirements is called Demand Management. Some industrial processes e.g. water pumping in water storage systems, aeration in wastewater treatment plants, industrial heating and refrigeration have an inherent flexibility within their existing assets [10]. This flexibility provides required storage to be used in active power manipulation tasks such as Frequency Regulation or so called Automatic Generation Control (AGC). These processes have enough power capacity, energy storage and no adverse reactive effect on the grid that make them suitable resources for grid balance applications. Their cycle efficiency is also very close to 100%. These loads can be used to provide real-time services to power system while fulfilling their primary obligation.

This paper deals with using flexibility of distributed responsive loads (DRLs) to minimize efficiency loss in a hydroelectric power plant. Flexibility of DRLs is modeled as regulation capacity to be used in grid balance service. DRLs are dispatched along with hydroelectric plants to increase generation efficiency without compromising their primary objectives. The objective function to be minimized is generation loss in the hydroelectric power plants. Minimization of generation loss is equivalent to maximization of generation efficiency. DRLs are treated as virtual power plants (VPP) and are distributed throughout the transmission system. Characteristics of power plants, DRLs and power system are modeled as constraints in optimization problem. These constraints are as follows:

• Power plant constraints: minimum and maximum power rating, ramp rate
• Responsive loads’ constraints: minimum and maximum power rating, storage and ramp rate
• System constraints: generation/demand balance

Two case studies are done to evaluate the effect of adding DRLs to regulation assets. The first case is dispatching DRLs along with all hydroelectric power plants in the system. In doing so, the operating points of the hydroelectric units move to the most efficient points on their efficiency curves. In the second case, only the swing plant which is operated in AGC mode is considered in objective function.

Results show the effectiveness of the proposed methodology in improving generation efficiency. The studied power system has 3 multiunit power plants. Daily “average produced energy to discharged water” is used as comparison quantity. The first case showed up to 2.25% of savings for 150MW of up and down regulation. The savings in second case is up to 0.2% since only one plant is considered. Sensitivity analysis shows that most of marginal savings are in the lower penetration of responsive loads.