A Generalized Formulation of Demand Response under Market Environments

Minh Y Nguyen* and Duc M. Nguyen

A Generalized Formulation of Demand Response

under Market Environments

Abstract: This paper presents a generalized formulation

of Demand Response (DR) under deregulated electricity

markets. The problem is scheduling and controls the

consumption of electrical loads according to the market

price to minimize the energy cost over a day. Taking into

account the modeling of customers’ comfort (i.e., prefer￾ence), the formulation can be applied to various types of

loads including what was traditionally classified as criti￾cal loads (e.g., air conditioning, lights). The proposed DR

scheme is based on Dynamic Programming (DP) frame￾work and solved by DP backward algorithm in which the

stochastic optimization is used to treat the uncertainty, if

any occurred in the problem. The proposed formulation is

examined with the DR problem of different loads, includ￾ing Heat Ventilation and Air Conditioning (HVAC),

Electric Vehicles (EVs) and a newly DR on the water

supply systems of commercial buildings. The result of

simulation shows significant saving can be achieved in

comparison with their traditional (On/Off) scheme.

Keywords: demand response, heat ventilation air condi￾tioning, electric vehicle, water supply systems, dynamic

programming, real-time markets, Smartgrid

DOI 10.1515/ijeeps-2014-0147

1 Introduction

Basically, Demand Response (DR) refers to the ability to

curtail some electrical loads, i.e., turn off appliances at

peak times to alleviate the need for peaking generation

sources. Thanks to Smartgrid technologies, e.g., Advanced

Metering Infrastructure (AMI) which enables the real-time

exchanges of information in the power system and makes

DR an important, growing part of resources for System

Operator (SO) to manage the balance between supply and

demand [1]. Under market environments, the benefits of DR

are achieved by customers’ intentional response to the

change of the electricity price over time; this action also

results in a saving of their electric bills. With flexible time of

response, a variety of ancillary services can be delivered

through DR such as load-frequency control, regulation,

spinning and non-spinning reserves, etc. [2].

Traditionally, electrical loads fall into three categories:

(1) sheddable loads, loads that, when turning on/off would

be slightly (or un-) noticed by the users (e.g., water heater);

(2) shiftable loads, loads with a flexible starting time that

only needs to be accomplished by a specified time (e.g.,

washing machine); and (3) critical loads, loads that, when

turning on/off could result in some inconvenience or dis￾comfort for the users (e.g., lights, air-conditioning.)

The first two categories, also known as Interruptible

Loads (ILs), are largely emphasized by the traditional DR

and considered in both wholesale and retail electricity mar￾kets. The scheduling of ILs, i.e., the portion and duration of

interruptions, as a part of Optimal Power Flow (OPF) pro￾blems is presented in Ref. [3]. Further in Ref. [4], it introduces

multi-objective optimization to take into account the con￾flicting targets, such as minimizing payment to ILs, minimiz￾ing frequency of interruptions, etc. at the same time meeting

the system requirements. In Ref. [5], DR is considered upon

the thermo-appliances of an aggregate load to form a Virtual

Power Plant (VPP); thereafter, the algorithm for short-term

operations in electricity markets is provided. In Ref. [6],

charging algorithm, namely smart charger for Electric

Vehicles (EVs) according to the price signal in Smargrids is

proposed; it is concerned with the trade-off between the

customers’ willingness to pay and the charging rate.

The third group, i.e., critical loads, gets attention of

modern DR: based on the physical model of loads, the energy

consumption can be managed to reduce the cost over a

period of time while maintain an adequate comfort level to

the customers. In Ref. [7], a real-time control for smart home

appliances in response to the real-time price is presented;

this combines an hour-ahead rolling optimization and real￾time control strategy to deal with complex operating envir￾onments. In Ref. [8], the stochastic optimization is used to

carry out the optimal operation of a building with both heat

and electricity consumption supplied by various energy

sources, including utility grid, battery, combined heat and

*Corresponding author: Minh Y Nguyen, Department of Electrical

and Computer Enigneering, Faculty of International Training, Thai

Nguyen University of Technology, 3–2 Str. Tich Luong, Thai Nguyen,

Viet Nam, E-mail: [email protected]

Duc M. Nguyen, International School of Education, Viet Nam

Maritime University, Lach Tray Str., Ngo Quyen, Hai Phong, Viet

Nam, E-mail: [email protected]

Int. J. Emerg. Electr. Power Syst. 2015; 16(3): 217–224

Brought to you by | South China University of Technology

Authenticated

Download Date | 6/3/15 4:35 AM