Measures to Improve Public Acceptance of Smart Metering System Identifying Requirements for Residential Consumers Rani Yesudas and Roger Clarke College of Engineering and Computer Science The Australian National University Canberra, Australia rani.yesudas@anu.edu.au, roger.clarke@anu.edu.au Abstract— Understanding the stakeholder’s needs, particularly battling to convince consumers of the potential benefits and the end-user’s need is important when designing and developing this may continue for years. A technology will not be wel- a critical infrastructure like power grid. Smart metering systems comed by the end user if it is not useful for them, even if it are considered as a vital element in grid modernisation projects. could contribute to solving major issues like lowering carbon It provides the utility with a range of opportunities to improve emissions and climate change [5] . their business. The benefits to other market parties are also clear. But the residential consumers are left behind. Many smart meter- The utilities introduced smart meters with the expectation ing projects, across the world, are facing consumers’ resistance. that energy consumers would use it as a tool to reduce peak Consumers perceive smart meters as an infringement to their energy usage. The concerned entities expected that, with few interests and rights. To avoid such situations, objectives of smart market choices, and a smart meter, the consumer would be well metering systems should also reflect consumer needs. Measures equipped to manage their consumptions efficiently [6, 7]. In need to be devised to elicit and include their requirements. To fact, most consumers feared that they would not be able to address this issue we analyse the reported concerns from the avoid the peak periods and that their bills would increase. They consumer and thereby identifying requirements for different suspected that the utility’s motive for the smart-meter rollout consumer segments. That enables us to propose functionalities and applications that will help the user utilise energy efficiently. was to make a profit at the cost of the consumer [8]. Further, when the system lacked visible benefits, but showed possibili- Index Terms—energy consumer, smart meter, smart metering ties for harm, other perceived risks like health and privacy system, advanced metering infrastructure, AMI, requirement became more prominent. elicitation, consumer needs. AMI’s business-centric characteristics lacks functionalities useful for the end-user and this is the main motivation for the I. INTRODUCTION the research described in this paper. The problem domain in AMI, related to the residential consumer, is relatively new in Electricity providers are facing challenge in implementing energy industry. In the traditional grid, consumers were passive new technologies to modernise the grid. Though the compo- users and they just had to pay for their usage. In the modern nents of the grid vary from region to region, the key elements grid, consumers are expected to become active members by are the same. Smart meter (SM) and the Advanced Metering managing their power usage. All the concerns from electricity Infrastructure (AMI) are considered as vital elements that can consumers imply that the implemented system is either not play an important role in managing peak demand [1]. useful for them or they have been ignored during requirement The energy industry is currently facing challenges in main- analysis. There is a possibility that the consumer requirements taining constant delivery of electricity to consumers. One of the were merely assumed by the analysts. Creating a successful main concerns with the power grid is in meeting the peak de- system also requires translating the end-users needs into the mand using the traditional infrastructure. Other issues include product scope. Late corrections of requirements errors are ex- the smooth integration of a range of low carbon technologies pensive and hence it is necessary to analyse and refine re- such as renewable energy sources and electric vehicles. A quirements before implementing a system. smarter power grid is seen as a necessity for an effective sys- Through this research we intend to understand consumer tem that is stable, reliable and secure [1-3]. The deployment of concerns and identify measures that could be applied to AMI smart meters on consumer premises has been the utilities’ start- projects to make them beneficial to the consumers as well. We ing point for grid modernisation. Such projects involve expen- have used Design Science [9, 10] as the research method for sive infrastructure that is paid for, whether directly or indirect- creating artifacts that embody such remedies. Buckminster ly, by consumers, and hence it is important to achieve consum- Fuller’s vision [11] is applied to analyse the problems that the er confidence. smart metering projects pose to the consumer. There have been multiple instances of major consumer Through this paper we try to identify consumer-focused so- pushback against smart meters, for example in Victoria, Cali- lutions using smart metering system. The remainder of this fornia and Ontario [4]. This has resulted in project-sponsors Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. paper is structured as follows. In Section II we provide a back- followed by subjective control were the main factors that af- ground on smart metering systems and in Section III we pro- fected a person’s attitude to use a smart meter. Perceived ease vide a literature review on residential consumer and their ener- of use affected perceived usefulness. [15] gy choices. In section IV we identify problems underlying the Stragier et al. investigated consumers’ perception of smart consumer concern. In section V we conduct a detailed problem meter and smart appliances in Flanders, Belgium. A large-scale analysis. In section VI we list the alternative measures based face-to-face user survey sought insight into the willingness of on existing technology. In sections VII and VIII we discuss the consumers to adopt smart meters. Factors included housing proposed measures and evaluate it. Finally in Section IX, we parameters, mobility, insulation measures, heating, lighting, conclude our work discussing future directions. energy patterns, domestic appliances, ICT and multimedia, ecological behaviour, ecological attitude, socio-demographic II. BACKGROUND parameters like gender, age, income and impressions of smart The main elements of the Smart Metering Sys- appliances. Perceived Ease of Use and Perceived Usefulness tem/Advanced Metering Infrastructure (AMI) are, the smart had significant influence on attitude towards smart metering meter, the head-end data storage within the utility and a net- system. Perceived Ease of Use had a strong influence on Per- work to support the two-way communication. The smart meter ceived Usefulness. [16] is designed to record detailed energy usage on the consumer Dunstan et al. investigated barriers of Demand Manage- premises. This detailed data is transmitted to the utility to ena- ment in Australia. About 808 groups from a range of stake- ble billing and Demand Response (DR) operation. holder groups participated in a survey. The statements “B22 - Federal Energy Regulatory Commission (FERC) defines electricity consumer lack interest in saving energy” and “B23 - DR as “the change in electric use by consumer from their nor- consumers want to use power when and how they choose” mal consumption patterns in response to changes in the price of were not thought by participants to be true. They thought that electricity and it also refers to the incentive payments designed consumers are interested in saving energy and that consumers’ to induce lower electricity use at times of high wholesale mar- reluctance to accept smart meters was because of a lack of ket prices or when system reliability is jeopardised”. Direct useful functionality [17]. Load Control (DLC) and Time-of-Use (TOU) pricing are the Firth et al. identified the influence of appliances usage and two main programs under DR. DLC is defined as a “mecha- energy user groups on demand in UK residential consumers. 72 nism by which the program sponsor remotely shuts down or dwellings at five sites were monitored over a period of 2 years. cycles a customer’s electrical on short notice”. TOU is defined Consumers were segmented into low, medium and high usage. as “a rate where usage unit prices vary by time period, and Appliances were classified as active, stand-by and cold use. where the time periods are typically longer than one hour with- Consumption increased more than 4% from the first to the in a 24-hour day. Time-of-use rates reflect the average cost of second year, with standby appliances being the major contribu- generating and delivering power during those time periods” tor, followed by active appliances. Low and high users in- [12]. creased consumption more than medium-usage consumers [18]. With TOU and DLC, utility can control customer’s energy McLoughlin et al. identified consumption characteristics usage. Utility providers expected that the consumer will use the based on dwelling and occupant characteristics in Irish dwell- TOU information to reduce electricity consumption during the ings. This study of 4200 consumers looked at the influence of peak period. Two ways in which the declaration is provided to appliances and of income on energy consumption. The largest the consumer are by means of an In-House-Display (IHD) contributors to demand were tumble dryers, dishwashers, elec- integrated with the smart meter, and a web-based ‘energy por- tric cookers and electric heaters as they all had large heating tal’. However, they are optional. components. The high energy consumers were mostly high Consumers have issues in accepting smart meter and its income professionals [19]. functionality. Opposition by residential consumers have be- come a big decision factor in some smart meter roll-outs [13, IV. KNOWLEDGE EXTRACTION 14]. Even a minority segment of consumers can become the A. Residential- Consumer Characteristics reason for abandonment of a project after research and devel- opment have been done and a great deal of funding has been Based on above consumer studies, the following can be committed. It is essential to understand consumer reaction to concluded. smart meters. 1. The primary factor that affects consumers’ acceptance of smart meters is its perceived usefulness. Ease of use of the III. LITERATURE REVIEW system significantly influenced usefulness. This is in In this section we analyse research studies that have been alignment with Technology Acceptance Model (TAM) conducted on energy consumers, their demand choices and factors; Perceived Usefulness (PU) and Perceived ease-of- factor affecting acceptance of smart meter. The articles that we use (PEOU) [20, 21]. have considered are listed below. 2. The second factor that affects consumers’ acceptance of Kranz et al. investigated smart meters’ acceptance factors in smart meters is subjective control. The consumer wants to Germany. An online survey on the project website attracted be in control of the operation of their appliances and power 351 participants in the age range 18 to 78. Perceived usefulness Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. usage. In addition they also want to be in control of the increased electricity bills associated with the use of smart me- data that is generated by smart meters. ters 3. Over time, energy usage increases for the same household 4) Safety Concerns living in the same conditions. This implies that consumers It has been reported that power surges have caused some are using more appliances that before. Having appliances smart meters to overheat and start a fire. Poor quality compo- in stand-by mode is also found to be a major contributor to nents and improper assembly of meters has been noted as the the problem. reason for overheating [24, 27]. 4. Energy wastage can happen due to carelessness and/or 5) Usability and functionality concerns ignorance. Most high energy users used carelessly as they By itself, a smart meter does not enable a consumer to could afford to pay for their usage. Low users who had manage their energy needs. They need both notification of been contributing to wastage did it mainly because of ig- peak-periods and the ability to turn particular appliances off or norance. They were not aware that certain choices were down. The possibility exists for home automation, if all appli- consuming power that could be avoided, e.g. leaving ap- ances could be operated via the smart meter, possibly remotely pliances on stand-by for long durations (several hours and as well as locally [28]. But these are aspirations rather than days) currently-delivered capabilities, and in any case it is not practi- cal for most residences, because people generally replace their B. Classifying consumer concerns appliances only at the end of their normal life In this section we are analysing the consumer concerns 6) Control and Choice concerns which will help in identifying solution. Information regarding There are also concerns about some consumers being una- system risks and consumer concerns were acquired, using pub- ble to avoid peak demand, resulting in huge electricity bills. lications by entities related to the power industry, and where Utilities expect consumers to utilise smart meter facilities, and necessary from media sources [20-25]. Based on the nature of adjust their consumption behaviour accordingly. On the other the consumer concerns, we classify them as follows: hand, consumers who stay at home most of the time can’t avoid • Health concerns peak as they will have to operate at least some and perhaps • Information security and privacy concerns most of their appliances. People with low financial capacity • Cost Concerns may not be able to invest in solar panels and other alternatives • Safety concern [28]. It is also difficult for most consumers to understand the • Usability and functionality concerns demand signals, decide on a course of action, and make the • Control and Choice concerns necessary changes. Consumers are also hesitant to give control 1) Health Concerns over their appliances to the utility. Consumers want to choose There have been complaints that smart meters cause health how they operate their appliances. problems. Complainants use the term Electromagnetic hyper- sensitivity (EHS) to refer to a range of health issues that they V. PROBLEM ANALYSIS link to transmitting devices. The existence of EHS is contested, Doubts exist about the basis of health concerns, and fire and some of the literature refers instead to Idiopathic (i.e. risk is a straightforward issue of product safety. This section ‘cause-unknown’) environmental intolerance (IEI) [22] . Smart considers the other concerns identified above, all of which are meters have a communication module for transferring data to related to the functionality of the smart metering system. the remote server, which may transmit and may receive using electromagnetic radiation. The meter manufacturers respond A. Factors Affecting Demand Choices that they have complied with industry standards, and that radia- Consumer resistance has often been associated with cir- tion from smart meters is much lower than that from mobile cumstances in which the demand response (DR) functionality phones [23, 24]. has been designed without taking into account important fac- 2) Information security and privacy concerns tors that affects consumer’s energy choices. Existing research The detailed data from smart meters are capable of reveal- [15-19, 29] suggests that following characteristics have been ing people’s lifestyle, occupancy and contents in a dwelling found to affect demand choices: [25]. Smart meter data is susceptible to modification and de- 1. Occupants in the dwelling: including sole- and multi- struction during transfer. There are also fears that blackouts and person occupancy, distinguishing families from non- other disasters could be caused by malicious hackers [26]. Util- related individuals sharing accommodation ities respond that the meters have a firewall and basic encryp- 2. Characteristics of the dwelling: particularly size, number tion and there have been no cases of reported attacks. of rooms, type (detached, semi-detached, apartment), con- 3) Cost concerns ditions (insulation, weather-proofness) The new infrastructure incurs huge cost. Some costs may 3. Occupants’ Propensity to Pay: particularly income and be borne directly by consumers, but in any case they fear that attitude towards convenience (balanced, anything for com- they will bear the cost of developing, running and maintaining fort, go-green, careless, enduring hardship due to no other the system. It is also not clear if smart metering system can alternative) reduce bills as claimed by utilities. Consumers have reported Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. Fig.1. TOU tariffs issued by Ontario Energy Board (OEB) effective from May 2015 [28, 29] 4. Appliance Working Mode: including stand-by, active,  Families with children fall into two segments: cold (appliance in continuous use but do not draw a con- stant amount of power). a) Working parents with school-going children can 5. Other factors: including climatic and weather conditions, avoid consumption during work-hours, but after and heath-related factors (e.g. continuous life support systems), before work-hours they have unavoidable energy re- security-related (e.g. security and theft control systems), quirements for cooking, cleaning, etc. It is not easy to and occupation-related (work at home office and produc- avoid those activities or the peak period 6 am – 9am tion units) and 4 pm – 7pm]. B. Scenario analysis b) Stay-at-home parents with young children can or- ganize some high energy consumption activities dur- The above analysis makes it evident all residential consum- ing off-peak time, e.g cooking and washing, but they ers cannot be assumed to be homogeneous. They do not all cannot avoid heating the dwelling, and if it is not en- respond to the demand response signal in the same way. In this ergy efficient their heating requirements will be high- section we analyze how different consumer segments will re- er. They can possibly go out during the day to reduce spond to the TOU signal on a winter day and the issues they energy usage, but it is not easy to make it a daily activ- will face. TOU Tariff rates effective in Ontario, Canada are ity. shown in figure 1. [30, 31]  Pensioners and people with medical needs face a situa- The higher rate is from 7 am to 7 pm, with peak-rates 7 am tion very similar to Stay-at-home parents, but as they may – 11 am and 5 pm – 7 pm. The analysis assumes that electricity have mobility issues and medical conditions that prevent is used for all purposes, which are primarily heating, cooking them from leaving out, they will have energy needs and cleaning. Then we identify several segments, referred to throughout the day. Moreover they may also have life- below as the common case and specific cases. supporting machines that will draw even more electricity. 1) The Common case  Work-from-home professionals work and live in the Most residential consumers are active between 7 am and 10 same premises, and hence have energy needs throughout pm. So they have some unavoidable energy needs during that the day. Their energy needs will be high throughout their period of the day. In general, heating will be required from work-hours. 5pm – 7 am in most cases and even more in others. From the above analysis it is evident that very few consum- 2) Specific Cases er-segments can adapt their energy needs to reflect the TOU  A single working professional is out of their dwelling pricing periods. In effect, TOU tariffs penalise many consumer during work hours. They can also avoid cooking by eating segments for factors that are outside their control. It is clear out. They can also choose to be away from home during that important end-user realities and needs have not been taken the peak demand periods. Their usage can then be limited into account during the requirement engineering process that to night/resting time. has been applied to the smart metering projects. Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. VI. ALTERNATIVE MEASURES TABLE I. USERS TYPES BASED ON FLEXIBILLTY WITH POWER USAGE Apart from accurate billing, the utilities’ main business goal in introducing smart metering system is to facilitate demand Flexible Consumers curtailment. Alternative measures could be considered which This category of users can follow the demand signals issued by may incur lower implementation cost and risks. Three such the utility. This includes mostly single, working, less stay at alternatives are discussed below: home consumers. Non-flexible Consumers A. Load shedding at feeder level Low usage but Low and Me- High usage – They Electricity substations have control mechanisms by which can’t avoid peak dium usage with have high energy load can be selectively shed at feeder level [32]. If the demand period - This constant usage requirements due to is much more than generation capacity, this mechanism can be category of con- throughout the various reasons. used to shed load in that area for short intervals. This demand sumer has low day - This category The high usage curtailment method is easier, as it does not have too many extra energy consump- of user includes consumers are dis- parameters to consider. The provider only needs to choose the tion but they can- stay-home parents cussed below. feeder and time-period for which the load has to be shed. This not avoid the with young chil- mechanism is already in use and usually the time and details of peak period. This dren, and pension- when the load will be shed will be declared in advance so that includes working ers. They may be the end-users can take necessary measures to overcome the families with able to shift some power loss. It can, however, have serious consequences for dependent kids. or a lot of their some consumer segments, such as people with medical needs high energy needs and home-working professionals. to off-peak or B. Higher tariffs for excessive usage beyond units allotted shoulder period. Another mechanism that is used for demand management is higher tariffs for excess usage. There is an allotted quantum of High usage – Non Flexible consumers energy for each household in a given period. Usage above the Due to unavoida- Due to work Due to high allotted amount incurs a higher tariff. This way demand is not ble medical con- from home/ home profile lifestyle curtailed, but the higher production costs can be transferred to ditions e.g. hav- based business - supported by very the user. ing life support These consumers high income - machines, some need energy supply These consumers C. Prepayment meters with restriction in the units allotted for consumers need for running their have very high a family to run various business without energy needs to Prepayment meters are in use in few countries. Payment has devices and hence any hindrance match their stand- to be made in advance, and the consumer can only use for what have less control during the business ard of living. They has been paid for. In countries where a consumer cannot be over demand. hours and hence may have an expen- disconnected by law, the end-user continues to get minimum They will need it cannot follow the sive security system amount of energy until the next payment is made. This method to be functioning demand signals. operating continu- could be used for limiting excessive usage. Every household 24x7. Some users may ously, lighting sys- could be allotted a set amount of energy, beyond which they only have medium tems and many cannot purchase, thereby restricting demand. energy requirement other devices. for their business. VII. PROPOSED MEASURES B. Identifying User Requirements A. Identifying User groups Based on the analysis done in the previous section, we pro- As it is observed that there are cheaper mechanisms to in- pose an informative feedback and varied billing for each con- duce demand reduction, the introduction of smart metering sumer types. Consumers need a system that will provide them systems requires justification. User requirements analysis needs with informative feedback on their usage patterns and provide to take into account not just the providers’ needs but also those them with hints to make intelligent choices and avoid energy of consumers. In this section using the artifacts identified by wastage. The customised billing options will provide user with analysing the problem; we are proposing consumer classifica- opportunity to make changes/ reduction in their usage pattern tion that will cover a wide variety of consumer and then identi- within their limitation. Table II and III provides user require- fy requirements that suit those segments. Consumer segmenta- ments for feedback and billing. tion is based on users’ capability to utilise the demand signals issued by the utility. The proposed classification is shown in Table I. Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. TABLE II. INFORMATIVE FEEDBACK FOR INTELLIGENT CHOICES settings. But as alternatives sumer off-peak rate 1. Their current usage feedback, they like solar pan- but it shouldn’t be as should be given el to supple- high as consumer 2. Energy consumption choices sufficient reminders ment their peak-rates, so that the  Saving mode – This mode should show them the most to alter their energy energy needs. consumer will be not cheapest energy choices choices. penalized for setting  Essential Usage mode – This mode should show them the up a home-based average required energy needs for the number of occupants business. They should mentioned. also be offered other  Maximum Usage mode – This mode should show them alternatives like solar maximum consumption limits under recommend standards panels to supplement for the total number of occupants so that the consumer un- their energy needs. derstands that beyond this level there is wastage occurring. 3. Energy saving tips - The system should include hints of items that could be operated/ avoided at different time of the C. Recommendations on System Modification day. It should be based on the demand signals from the utility. On the basis of the analysis conducted above, it is now fea- This will exclude the guess work out of the user sible to propose modifications to existing smart metering archi- 4. Warnings for high usage - Warning should be provided dur- tecture. Currently, in the smart metering system there is only an ing i) peak period and ii) Over usage during a billing cycle. In-House-Display (IHD) and an online energy consumption 5. Provision to set consumption limits – There should be option displaying system for feedback. The IHD displays the con- for the consumer to provide a desired bill amount for the billing sumption data recorded in the consumer’s smart meter and the period and get information on consumption limits within that online system displays the energy data that is transmitted and billing cycle. The information should show the user how much stored at the utility’s side. Both the systems also display the energy they can use per day to match the billing amount. demand signal like peak period and tariffs rates. The need ex- ists for an intelligent feedback mechanism rather than just dis- TABLE III. BILLING OPTIONS TO SUIT CONSUMER CATEGORY play of usage data and tariffs. The current data does not make the consumers well informed about the choices they have. Billing based on TOU Hence the smart metering system needs to incorporate a system This aligns with utility providers’ current plan to charge user that uses both consumer input and signals from the utility. As based on TOU rates. the system provides the user options to input information on Billing based on Usage energy behaviour, it is referred to as Consumer Energy man- This is calculated based on number of occupants and other agement System (CEMS). Additional information such as parameters weather data may also be drawn from external entities for fine Usage within a recom- Usage above a recommended tuning the calculations. An overview of the modified system is mended level – People using level - these users can be fur- provided in figure 2. The recommendation on billing options within recommended limits ther classified into three: high can be done within the smart metering module based on the and with minimal wastage of profile and careless users; setting provided by the utility. energy shouldn’t be penal- people with medical conditions ized for not being able to and home based business. The D. Proposed Functional requirements avoid the peak period. They users with usage above rec- 1) Accurate Usage details. should have an option to pay ommended level is further This feature is currently present in many smart meters. The lower rates even for the peak discussed below. usage is displayed using the IHD integral to or connected with period. This will encourage the smart meter and also in the provider’s online usage display them to continue using less system. This usage data may also be displayable in the pro- energy. posed CEMS by synchronizing it with head-end data. 2) Energy consumption and Billing related advices. These advices mainly rely on consumer input. CEMS will Usage above a recommended level allow the user to input data related to dwelling characteristics, Careless and High- People who People doing busi- occupant characteristics and appliance characteristics. This will profile users - They can’t avoid ness/work at home – give sufficient information to calculate the energy requirement don’t require any due to various They should have for the household. It can be fine-tuned with the weather data concessions on their limitations - different tariff rates from external entities. The accuracy of the calculation will be bills as their high They require for business hour and dependent on the information provided by the user. Further, the usage is their concessions in after business hours. system will have to fetch the demand signals from the utility. choice. They have billing and Rates for commercial Based on this information, the billing and warning information to pay as per the they should be purposes should be can be generated. To calculate consumption information by utility provider’s offered other higher than the con- providing a bill amount, the demand signal from the utility alone is sufficient. The user can input a desired amount and the Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. Fig.2. Proposed Smart metering system architecture TABLE IV. PROPOSED MODE OF OPERATION FOR CEMS system can calculate the consumption threshold to match Mode Operating pattern the desired bill amount. Standalone The application should be able to work 3) Billing Option for different Category without connecting to the rest of the system. The smart meter is capable of storing accumulated usage The present demand signals can also be fetched data for each demand period (peak, off-peak and shoulder). The from the utility’s website and entered manually. rates that are applied will vary according to the user type. The Consumers who are particularly concerned head-end will have to store a user profile based on the user about privacy and data-sharing with the utility category and it will have data regarding allotted energy for the can use the application in this mode. There household, the category the household belongs to, and the rates should also be option for these consumers to that apply for the category. For each billing cycle, this infor- synchronize their offline details with the utility mation will be applied to generate a customized bill for the if they wish to. To run the application in this user. For a home-based business, because their power-usage for mode, the user does not even need a smart me- commercial purposes depends on their business-hours, it is ter in their premises. better to have their meters accumulate data based on work-time (working hours, outside working hours) rather than demand Online and The application should be able to work with period. partially minimal sharing options. This option is useful synchronize for user who would like to get favorable billing E. Proposed Non-Functional requirements with utility option from the utility. The user can share their 1) CEMS should be able to operate both online and as a self-entered user-profile with the utility. They standalone program. can also opt for receiving the demand signals The proposed system should be able to operate in 3 differ- when connected online. Even for this mode, the ent modes as stated in Table IV. user does not even need a smart meter in their 2) User profile data stored by the utility is used only in premises. accordance with the terms of the consumer’s consent Online and The application should also be able to work A user profile will be stored at head-end to choose the bill- fully syn- in full automation. In this mode the system will ing category and other related matter. The utility should ensure chronize fetch consumer’s usage data stored in the head- that the details are only used for the billing purpose and that a with utility end, receive all alters and demand signals from third party can gain access to it only with consumer consent. the utility and provide all possible hints and tips to the user. For the application to operate in this mode the user requires a smart meter and they should consent the utility to have remote access to their data. Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. VIII. SYSTEM EVALUATION B. Effectiveness of the system In this section we discuss the effectiveness of the smart me- A. Evaluation of Design tering system in a particular setup. For that we choose the smart This section discusses how the proposed features affect the metering project in Australia. The Department of the Envi- current system. ronment, Water, Heritage and the Arts (DEWHA) prepared the 1) Feasibility with the current context initiatives for the smart grid system in Australia and it primari- The design measures that we propose primarily add end- ly focused on distribution and retail value chain elements. Two user value to the smart metering system so that consumers can main applications that were considered as part of the primary make informed choices. Current smart metering systems have smart grid technologies were the customer-side application and an online energy management system that displays the con- its key enabling application. sumer’s usage, alerts from the utility and demand signals. From DEWHA also insisted that trials should be made to fully this information, users have to make assumptions and make test the benefits of the consumer applications. For the trail they decisions. With the CEMS, the information is already pro- also suggested that customer segments with different patterns cessed so that it can be easily used by the consumer. Even of electricity use be identified and included, because each seg- consumers who have opted out of smart meters installed in ment appeared likely to respond differently to the customer their premises can use this system because it can work as a applications components being tested [33]. The Victorian standalone tool. Even if the user does not have access to real Government mandated the implementation of smart meters for time consumption data, they can receive tips on how to reduce residential customers in 2006 and the rollout commenced in energy wastage based on the data they input to the system. 2009. The system mostly used RF mesh technology for its 2) Effects on other Smart Metering Functionality communication. Energy price deregulation was also introduced The proposed measures do not affect the technical specifi- for all consumers to complement the smart meter roll-out. But cation of the smart meter. The billing categorization is a cus- there was huge resistance the TOU tariff scheme from consum- tomizable feature. Smart meters have the capability to catego- ers and advocacy groups such that the providers had to permit rize usage data based on time period. The time period can be consumers to remain on the old flat-rate tariff, even after the programed to the meter to classify consumption. Currently this deployment of smart meters [34]. is done based on demand period, because this setting is suffi- The Australian Smart meter scheme clearly shows that cient for most of the user categories. For home-based workers, smart meters were rejected by the consumers as it was not clear the meter needs to be programmed to classify time period how the consumers could benefit from the system. Our pro- based on work-time. These are customizable settings within the posal considers different consumer segments and suggestions smart meter that can be easily changed. These features also do that support each group. Using our proposed scheme different not affect remote operations or communication features of the user groups can be provided different billing options. smart meter. 3) System Limitations IX. DISCUSSION AND CONCLUSION The two design adaptations proposed are dependent on user Smart metering systems have the potential to contribute to information. The user profile needs to be accurate as it will be smart grids and improved energy management. These systems used to calculate bills. Hence the utility will have to use other benefit the utility because billing is more accurate and manual verification methods to make sure that such information pro- reading costs and errors are avoided. The infrastructure re- vided by the user is truthful. They will need to verify tenancy quired for the system is vast and the meters are expensive. This contracts, income certificates and government approved docu- cost is ultimately borne by the consumer, yet in projects to date ments on special needs and family member listings. they have gained very limited benefits from smart meters. For the feedback system, the data accuracy is not as vital as To tackle consumer concerns the commonly provided solu- it is for the billing system. The users should be informed that tion is to educate and engage the consumer to use the the calculations will be based on their input and hence discrep- system[35, 36]. But that does not justify the need to introduce ancies will be observed if they don’t provide accurate infor- smart metering system. The alternative measures we have sug- mation. The utility provider can educate the consumer in ap- gested in section VI are sufficient to reduce demand. If con- propriate use of the facility. sumers have to actively engage in power management, the The details entered for the feedback system could profile system should provide them with useful and easy to use choic- the user. It could provide [16] detailed information about appli- es. ances, their use, and the lifestyle of the user. These recommen- As part of our research we have earlier identified a security dations therefore do not solve the existing privacy issues that analysis frame work specifically for the smart grid [37], options the system is facing, but neither does it aggravate it. But con- in smart metering system to improve privacy and provide more sumers have choice to work with the application as standalone control options to consumers [5]. We have also identified spe- and that way they can choose not to share their information cific functionalities to benefit different consumer groups [28]. with the utility. Needless to say, the system needs to be secure, In this paper we have identified different feedback and billing to ensure there is no theft of user profiles and user data. options to suit different consumer segments. A limitation to the research was that user survey and field testing was not a feasible. Moreover we have not considered Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. the usage of electronic vehicles (EV). EVs may draw more [11] H. Brown, R. Cook, and M. Gabel, Environmental power. Hence it can make a huge difference in the power re- Design Science Primer: Advocate Press, 1978. quirement for the concerned consumer. But we would like to [12] FERC, "Assessment of demand response and suggest keeping EV power requirements separate from other advanced metering," 2008. [Accessed 01 March residential energy requirements until they are widely used. It 2015]. would be preferable to keep a separate power line for charging [13] T. Bayar. (2013, Will Germany Reject Smart Meters? EVs and billed separately. Available: We started our analysis by choosing the problem situation http://www.renewableenergyworld.com/rea/news/arti in smart metering system. Then preferred states were identified cle/2013/09/will-germany-reject-smart-meters . by extracting consumer-friendly user requirements. Based on [Accessed 01 March 2015]. the requirements, an abstract architecture and functional speci- [14] A. Colley (2014, Half of users abandon smart meter fication were proposed. It was then shown how those features trial - Australia. Available: could be added to the existing system. We have proposed adap- http://www.itnews.com.au/News/370919,half-of- tations to the artifacts that feature intelligent feedback mecha- users-abandon-smart-meter-trial.aspx . [Accessed 01 nisms and more billing options to suit different consumer cate- March 2015] gories. The evaluation suggests that current smart metering [15] J. Kranz, J. V. Gallenkamp, and A. Picot, "Exploring systems can be readily modified to add the proposed features. the Role of Control-Smart Meter Acceptance of But the efficiency relies on the accurate details provided by the Residential Consumers," in AMCIS, 2010, p. 315. user. [16] J. Stragier, L. Hauttekeete, and L. De Marez, "Introducing Smart grids in residential contexts: Consumers' perception of Smart household REFERENCES appliances," presented at the Innovative Technologies for an Efficient and Reliable Electricity Supply [1] Y. Yan, Y. Qian, H. Sharif, and D. Tipper, "A Survey (CITRES), 2010. on Smart Grid Communication Infrastructures: [17] C. Dunstan, N. Ghiotto, and K. Ross, "Barrier to Motivations, Requirements and Challenges," Ieee demand management : report#2 of the Australian Communications Surveys and Tutorials, vol. 15, pp. Alliance to Save energy " Alliance to Save Energy, 5-20, 2013. Tech. Rep, 2011. [2] H. Farhangi, "The Path of the Smart Grid," Ieee [18] S. Firth, K. Lomas, A. Wright, and R. Wall, Power & Energy Magazine, vol. 8, pp. 18-28, Jan- "Identifying trends in the use of domestic appliances Feb 2010. from household electricity consumption [3] S. M. Amin and B. F. Wollenberg, "Toward a smart measurements," Energy and Buildings, vol. 40, pp. grid: power delivery for the 21st century," Power and 926-936, 2008. Energy Magazine, IEEE, vol. 3, pp. 34-41, 2005. [19] F. McLoughlin, A. Duffy, and M. Conlon, [4] UKERC, "Future Research Requirements for Smart "Characterising domestic electricity consumption Metering Workshop," The UK Energy Research patterns by dwelling and occupant socio-economic Centre2011. variables: An Irish case study," Energy and [5] R. Yesudas and R. Clarke, "Architecture and Data Buildings, vol. 48, pp. 240-248, May 2012. Flow Model for Consumer-Oriented Smart Meter [20] F. D. Davis, "Perceived usefulness, perceived ease of Design," in Information Systems Development: use, and user acceptance of information technology," Transforming Organisations and Society through MIS quarterly, pp. 319-340, 1989. Information Systems (ISD2014 Proceedings), [21] V. Venkatesh and F. D. Davis, "A theoretical Varaždin, Croatia, 2014. extension of the technology acceptance model: Four [6] T. J. Lui, W. Stirling, and H. O. Marcy, "Get smart," longitudinal field studies," Management science, vol. Power and Energy Magazine, IEEE, vol. 8, pp. 66-78, 46, pp. 186-204, 2000. 2010. [22] C. Baliatsas, I. Van Kamp, E. Lebret, and G. J. [7] M. Spindleruv, "How Smart metering pilot project Rubin, "Idiopathic environmental intolerance lost illusions and ideas or nothing is just Black and attributed to electromagnetic fields (IEI-EMF): a White!," Enel2012. systematic review of identifying criteria," BMC [8] G. Zachary, "Saving smart meters from a backlash," Public Health, vol. 12, p. 643, 2012. Spectrum, IEEE, vol. 48, pp. 8-8, 2011. [23] T. Krishnamurti, D. Schwartz, A. Davis, B. [9] S. T. March and G. F. Smith, "Design and natural Fischhoff, W. B. de Bruin, L. Lave, et al., "Preparing science research on information technology," for smart grid technologies: A behavioral decision Decision support systems, vol. 15, pp. 251-266, 1995. research approach to understanding consumer [10] A. R. Hevner , S. T. March, J. Park, and S. Ram, expectations about smart meters," Energy Policy, vol. "Design science in information systems research," 41, pp. 790-797, Feb 2012. MIS quarterly, vol. 28, pp. 75-105, 2004. Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes. [24] J. Betz (2010, Smart Meters Under Fire as Electric [33] DEWHA, "Smart Grid, Smart City: A new direction Bills Soar. Convention for the Protection of Human for a new energy era " Department of the Rights and Fundamental Freedoms as amended by Environment, Water, Heritage and the Arts 2009. Protocol No. 11. [Accessed 01 March 2015]. [Accessed 01 March 2015]. [25] E. L. Quinn, "Privacy and the new energy [34] J. Benvenuti, "Lessons from Victoria: Has infrastructure," Available at SSRN 1370731, 2009. deregulation delivered?," Consumer Utilities [26] T. Baumeister, "Adapting PKI for the Smart Grid," Advocacy Centre ( CUAC), Brisbane2013. 2011 Ieee International Conference on Smart Grid [35] Y. Strengers, "Smart metering demand management Communications (Smartgridcomm), 2011. programs: challenging the comfort and cleanliness [27] EMFSN, "Smart Meter Fires and Explosions," 2012. habitus of households," in Proceedings of the 20th [28] R. Yesudas and R. Clarke, "Identifying consumer Australasian Conference on Computer-Human requirements as an antidote to resistance to smart Interaction: Designing for Habitus and Habitat, 2008, meters," in Innovative Smart Grid Technologies pp. 9-16. Conference Europe (ISGT-Europe), 2014 IEEE PES, [36] N. Boughen, Z. Castro, and P. Ashworth, 2014, pp. 1-6. "Understanding the residential customer perspective [29] S. Darby, "Smart metering: what potential for to emerging electricity technologies: Informing the householder engagement?," Building Research & CSIRO Future Grid Forum," Brisbane, Queensland: Information, vol. 38, pp. 442-457, 2010. CSIRO, 2013. [30] OH. (2015). Time-of-Use Pricing. Available: [37] R. Yesudas and R. Clarke, "A Framework for Risk http://www.ontario- Analysis in Smart Grid," in Critical Information hydro.com/index.php?page=current_rates . [Accessed Infrastructures Security, ed: Springer, 2013, pp. 84- 01 March 2015]. 95. [31] OEB. (2015). Time-of-use (TOU) Prices. Available: http://www.ontarioenergyboard.ca/OEB/Consumers/ Electricity/Electricity+Prices. [Accessed 01 March 2015]. [32] S. Sridhar, A. Hahn, and M. Govindarasu, "Cyber– physical system security for the electric power grid," Proceedings of the IEEE, vol. 100, pp. 210-224, 2012. Copyright © 2015 for this paper by its authors. Copying permitted for private and academic purposes.