Development of a Demonstration Rig for Providing Primary Frequency Response through Smart Meters W.M.T.Vijayananda Ceylon Electricity Board, Sri Lanka gllps@sltnet.lk Kamalanath Samarakoon, Janaka Ekanayake School of Engineering, Cardiff University kamalanath@ieee.org, EkanayakeJ@Cardiff.ac.uk Abstract- A development of a rig that demonstrates a load control scheme which provides primary frequency response is reported. The rig uses a commercially available smart meter and remotely controlled smart sockets. A load controller implemented in a computer sends control signals to switch smart sockets on and off. The meter communicates with the computer using Modbus protocol through RS485 – RS232 protocol convertor. The controller sends load control signals to smart sockets via Home Area Network which communicates using ZWave protocol. Index Terms-- Demand Response, Demand Side Management, Home Area Network, Primary Frequency Response, Smart Grid, Smart Meter, Modbus, ZWave, I. INTRODUCTION Worldwide, smart meters are being introduced as a measure of increasing power system operational efficiency and also reducing energy consumption by providing accurate information to consumers [1]. The real time power consumption information will be provided to consumers through display devices together with time-of-use tariffs. It is expected that, based on the information, consumers would change their behaviour to reduce their electricity consumption or would shift usage of electrical appliances to lower tariff periods. In addition to providing information, it is expected that smart meters would support operation of the grid. In a power system, demand and generation should be balanced in real time. In an event of a sudden imbalance due to a loss of generation or connecting a large load, frequency starts to drop rapidly. During such an event, the governors of partially loaded generators increase their output to stabilise the frequency drop. The amount of generation that is available within first 10 s and sustained for another 20 s is called the primary frequency response in the UK [2]. In addition to the generators, primary frequency response is obtained through demand reduction. In the UK this service is called Frequency Control Demand Management (FCDM) service [3] which is secured through commercial arrangements. The FCDM service switches off large loads by using under frequency load shedding relays typically set at 49.7 Hz. The use of local frequency measurements to provide primary frequency response by controlling loads is reported in the literature [5-10]. These proposals use standalone controllers or controllers fitted onto electrical appliances to responds to frequency deviations. The initiatives of smart meter deployments offer new opportunity to use them for primary frequency response. A load control scheme that uses smart meters to provide primary frequency response by local frequency measurements has been reported in [11]. In this paper a control algorithm which was developed for load control scheme, was implemented in a laboratory rig and results are reported. II. LOAD CONTROL SCHEME In the load control scheme reported in [11], household electrical appliances are plugged into remote controlled plugs which are piggy backed onto existing wall sockets. The remote controlled plugs are switched off in order to reduce the demand thus providing primary frequency response. In the algorithm which was developed for the load control scheme, domestic appliances are categorised into five groups. They are named as NCC (Fridges, Freezers and Space heaters)), NCS (Washing Machines and Dryers), NCE (Ovens and Hobs), E (Inline Heaters) and C (Lighting) depending on their impact on customers’ comfort and safety. The algorithm reads power system frequency from the smart meter and switch off different appliance groups depending on the frequency drop. As shown in Fig, 1, the load control scheme defines switching off (F OFF ) and on (F ON ) frequencies and operating times for each group. The operating times are Load Switch Off Period (T O ), Frequency Monitoring Period (T M ), Delay Period (T D ) and Maximum Load Reconnecting Period (T R ). 50 C Start of the event 49.8 E 49.7 48.8 48.9 49.5 Load Shedding 49.0 49.2 Minimum normal frequency Statutory frequency limit Load shedding frequency 0 5 10 15 30 40 50 60 70 80 90 100 110 120 Time (s) Primary Frequency Response Secondary Frequency Response Load Switch Off Period Maximum Load reconnecting period NCE NCC NCS Typical FCDM relay setting Frequency Monitoring Period TO TR TM TM TM TR T O T O TO Notes: Different load categories have different time periods Time axis has two different scales TR TR TR Delay Period TD F NCE ON TD TD F NCC ON F NCS ON Up to 30 minutes TO F NCE OFF F NCC OFF F NCS OFF F E OFF F C OFF Fig 1. Timing of the load control algorithm