Feedback and other measures targeting user behaviour

Feedback measures aim to increase end users’ awareness and transparency about their own levels of energy consumption. Combined with practical tips for energy-efficient user behaviour and its potential energy and cost savings, feedback and other measures can motivate users to change their behaviour, but also to invest in energy saving technologies. Feedback measures cover a range of practices, including individual metering, smart metering with feedback on tips where to save, and informative/comparative billing as well as prompts, motivation campaigns for behaviour change with practical tips for energy-efficient user behaviour and its potential energy and cost savings, energy saving competitions, and training for users (mainly in the service sector).

Worldwide implementation status
In recent decades, governments have increasingly recognized the importance for consumers’ energy consumption behaviour and required the delivery of more effective feedback to consumers.
For example, in the European Union (EU), the European Commission and EU member states are working together for the roll-out of smart meters and aim to equip at least 80% of users with smart meters by 2020 (EurActiv 2012, September 18).
At the country level, for instance, in Denmark, there is a legal obligation to provide an “informative electricity bill” indicating environmental impact as well as historic and normative comparisons (Fischer 2007). In Italy, smart meters have been deployed since 2001 by the energy supplier Enel. Based on the evaluation of that deployment, the Italian government started the mandatory roll-out as early as 2008 and has provided financial incentives to network operators, due to the large cost they bear.
In addition, energy companies have also launched contests to encourage customers in energy saving, and recently also through the use of smart meters. For example, the ‘Biggest Energy Saver Campaign’, which has been organized annually by the Texas-based utilities and technology companies since 2011, is designed to reward consumers who make the most of their smart meter data to reduce their energy consumption (http://www.biggestenergysaver.com/).

Governance level
Feedback and behavioural measures have been implemented at trans-national, national, regional, and local level all over the world.

Related sectors

• services
• residential
• industrial

• Refrigerators and freezers
• Clothes Washers
• Clothes Dryers
• Kitchen stoves and ovens 
• Coffee machine
• Microwave
• Dishwasher
• Home Audio 
• Television
• Monitors
• Computer
• Office equipment (Copiers, Fax machines, Printer, Server etc.)
• Data centres
• Commercial Food Cooling Equipment
• Commercial Laundry equipment

Related target groups

• User

Direct beneficiaries

• User

  Users obtain a better understanding of their energy consumption and the options they can do to save energy and costs, and, in case of deploying the smart metering system, a potential cost reduction due to energy use shifted to off-peak time as indicated by the meter.

Indirect beneficiaries

• Manufacturer

  Manufacturers of energy efficient appliances can benefit from the feedback measures indirectly, because of the potential demand increase of their products.

• Energy company

  For energy companies, smart meters especially offer opportunities to optimise the matching of energy supply and demand, particularly with respect to increasing shares of fluctuating renewable energy generation in the network, and to save costs on metering and billing.

• Governmental actor

  Energy regulators can indirectly benefit from the roll-out of smart meters that enable a better energy supply regulation, i.e. reliability and voltage quality (Balmert et al. 2012).

Relevant barriers

• Knowledge and information barriers: The main barrier addressed by these type of policies and measures is that end-users are not informed enough about the exact energy consumption of their appliances. Feedback measures will increase end users’ awareness and transparency about their own levels of energy consumption
• Lack of interest and motivation in energy efficiency-improvement: There is a tendency that households focus on their core activities only. If the end-user is convinced that the energy consumption of appliances is very high this could be a motivation to invest in energy-efficient appliances and to change the behaviour.

Relevant incentives

• Saved energy costs: After a period of time, the upfront costs, which are normally higher for energy-efficient appliances are amortised due to lower running costs. End-users can save energy costs if they reduce their stand-by losses.
• Contribution to protection of the environment: This may be an intrinsic motivation for any actor.

Feedback and behaviour measures already include different instruments, i.e. individual metering and billing as well as the provision of information and training to users about how to operate and benefit from the meter use, demand response, energy efficiency options, etc.
The package works more effectively for encouraging users’ energy saving behaviour and uptaking of energy efficiency technologies, when it is combined with other instruments, such as:

• Energy advice & assistance: it can facilitate the uptaking of energy efficiency options after users realize their energy consumption and saving potentials with feedback measures. 
• Energy labels: if the tips involve replacement of an old appliance, these will be instrumental for selecting an energy-efficient new one.
• Regulations: a proper regulatory framework should be in place to support the roll-out of individual and if possible smart meters or informative billing, for example, by setting obligations for responsible actors (e.g. network operators), addressing privacy protection and protecting consumers from unjustified energy rate increases (van Elburg 2011; Fischer 2007; Balmert et al. 2012). Besides, if utilities are obligated to conduct energy efficiency programmes, the government can recognize feedback measures as part of the programme, if the savings can be proven (Duscha & Dünnhoff 2007). In addition, in terms of smart meters, widely accepted standards of hardware and software, communication protocol, and data formats should be developed. Meanwhile, national technical standards and regulations for metering and measurement need to be reviewed (Balmert et al. 2012). 
• Financing scheme: as part of finding solutions of cost-sharing among different beneficiaries for the roll-out of smart meters or informative billing, the government needs to provide appropriate regulatory or financial incentives to responsible actors (e.g. network operators) to cover their costs.

The following pre-conditions are necessary to implement Feedback Measures targeting user behaviour:

Agencies or other actors responsible for implementation
For feedback and behavioural measures in general, energy companies are needed for the metering and the feedback and potentially the advice on energy-efficient behaviour. The latter may also be provided by energy agencies or consumer information centres.
In terms of smart meters, since its roll-out involves various actors, such as users, network operators, energy supply companies, manufacturers of smart meter devices and corresponding government agencies, the establishment of a cross-sectoral committee that co-ordinates activities of all these actors may facilitate the roll-out process.
The agency and authorized organisations that are responsible for testing and evaluating smart meters should be in place to ensure their quality.

Funding
As part of finding solutions towards cost-sharing among different beneficiaries for the roll-out of smart meters or informative billing, the government needs to provide appropriate regulatory or financial incentives to responsible actors (e.g. network operators, utility companies) to cover their costs

As with all measures, but particularly to roll out new feedback technologies such as smart meters, a cost-benefit analysis should first be conducted in order to assess the cost-effectiveness of the large-scale deployment of this technology. In addition, given the social, economic, and health concerns from general public, communication activities should be carefully planned before the roll-out. Furthermore, pilot projects are essential for the roll-out, due to the existence of various socio-economic and/or technical uncertainties.

Quantified target
Feedback and behaviour measures do not usually have a quantified target in terms of energy saving, but they may specify the numbers of participants and units, for example, 80% of consumers being equipped with smart meters and receiving advice.

International co-operations
Policy makers can benefit from international co-operations that share the roll-out experiences of feedback and behaviour measures. In addition, international co-operation on the development of smart metering system technologies can be helpful for overcoming the technical barriers and reducing costs.

Monitoring
Smart meters and informative billing are themselves tools that track energy consumption. For other feedback and behavioural measures, monitoring should include the number of consumers having received feedback and advice, and what the frequency and subjects were. Costs should also be monitored.

Evaluation
Evaluation can be conducted through controlled experiment, before-after comparison of energy consumption and users’ behaviour, etc., and a control group. In addition, the features of feedback measures, such as delivering frequency, content, the involvement of comparison (e.g. historical comparison, normative comparison, etc.), need to be included in the evaluation for further improvements of the design of feedback measures (Fischer 2007).

Relevant individual dynamics and side-effects

• Dynamic market transformation and/or innovation: Governments’ promotion of installing smart meters facilitates the market transformation of smart meter systems and devices. Indirectly, feedback measures might increase the demand for energy efficiency products and services as the users become aware of energy saving potentials of their appliances.
• Avoiding lost opportunities: Through feedback and behaviour measures, especially smart metering and informative billing, users are better able to discover their energy consumption and energy system than is possible with their conventional metering and billing system. Thus, the behavioural change and potential uptaking of energy efficiency technologies induced by the consumers’ discovery and follow-up learning avoids the loss of opportunities.
• Minimising rebound effects: In fact, feedback and other behavioural measures are an important means for tackling rebound effects that may occur after technical energy efficiency improvements. Such measures can educate users about the appropriate use of energy-efficient appliances. If applied long, well and frequently enough, feedback and other behaviour measures may motivate users to change their behaviour and develop energy saving habitat, which in turn minimizes the rebound effect.
• Minimising snap-back effects: The danger of snap-back effects is particularly high for this type of measure: energy savings will be lost, if users fall back to previous, less energy-efficient behaviour and routines. Feedback and behavioural measures regularly implemented over a certain period (at least three months) enable users to develop energy saving habitat and monitor the effects of their energy saving behaviour. Continued feedbacks and saving tips are needed to maintain the behaviour change and thus to avoid snap-back effects (Darby 2006). The regularity and continuity of feedbacks can be easily realized by smart meters.
• Minimising free-rider effects: Free-rider effects can occur with feedback and behaviour measures, if users would have implemented energy-efficiency options anyway. However, such effects are likely to be small. There will probably be little behavioural change and awareness of saving options without such measures.
• Creating Spill-over effects: Feedback and behavioural measures are likely to have positive spill-over effects. For example, consumers’ knowledge about energy consumption and tips for energy saving behaviour are likely to be spread by word-of-mouth and results in behavioural change of other users.

Design for sustainability aspects
Public concerns about the health risk accompanied with smart meters (http://stopsmartmeters.org/) and social issues such as privacy protection (van Elburg 2011) need to be well studied and communicated before its roll-out. In addition, feedback measures need to design tailored services to help low-income or low-education customers (Consumer Focus 2010).

Co-benefits
Both the increasing demand for smart meters and potentially the demand for energy efficiency products and services create job opportunities in these fields.

The following barriers are possible during the implementation of the policy

• Users are concerned about privacy protection, health risk, and potentially higher energy bills (Darby 2006; BEAMA et al. 2010; van Elburg 2011).
• Users may not be aware of a smart metering initiative and the benefits of its application (Darby 2006; BEAMA et al. 2010).
• The large-scale roll-out of smart meters can be very costly. The responsible actor (e.g. network operators) may not be willing to invest in the technology if the cost they bear cannot be accommodated by the revenues, for example, by increasing energy tariff to users. However, opposition to increasing energy tariff can be strong from both users and the government. Additionally, the inequitable share of the costs and benefits among different beneficiaries may also lower the incentive for the responsible actor (BEAMA et al. 2010; Balmert et al. 2012). 
• There is a lack of standards to specify the functionalities of smart meters and guarantee their interoperability and use for feedback (BEAMA et al. 2010)
• Technical barriers also exist, for example, the lack of the modularity and flexibility of existing smart meters, which makes it very costly for customising systems to meet the local requirements (BEAMA et al. 2010).
• There may be a shortage of skilled workers who are able to install the smart metering system (BEAMA et al. 2010).

Regarding more informative and frequent billing: 

• It may face resistance from the energy company, which has to bear the additional costs (Duscha & Dünnhoff 2007).
• Its design may not be effective enough to encourage users to change their behaviour

The following measures can be undertaken to overcome the barriers

• In order to strengthen users’ awareness of energy savings and their confidence on smart meters, communication and information provision need to be carefully planned and set up, where social and health concerns, energy tariff increase, and potential energy saving options should be clearly communicated. Possible methods include stakeholder meetings, consumer advisory group, consumer engagement campaign using social marketing approach, demonstration projects (Darby 2006; EurActiv 2012, June 21; Consumer Focus 2010; Balmert et al. 2012). 
• To ease users’ learning about their energy consumption, feedback illustration needs to be designed in a way that includes multiple options for feedbacks, interactive components, appliance-specific breakdown, comparisons with previous periods and similar households, ambient devices, and time-of-use prices (Fischer 2007; Darby 2006). 
• Regulations should be formulated to protect users’ data privacy, for example, allowing customers to decide which measurement data to be used by which party (van Elburg 2011), clearly stipulating how the gathering, processing, storage, and evaluation of data is used as well as who has access to which data (Balmert et al. 2012).
• The government should commit to setting up a justified and transparent cost-sharing framework, i.e. the costs borne by the network operator should be covered and only justified costs are passed through to users and other beneficiaries (Balmert et al. 2012). 
• Comprehensive cost-benefit analysis should be carried out to show whether the roll-out of smart meters is cost-effective and to assess the distribution of costs/benefits among various actors (EurActive 2012, June 21). In addition, social and environmental impacts should be also assessed (Giordano et al. 2012). 
• Widely accepted standards of hardware and software, communication protocol, and data formats should be developed (Balmert et al. 2012). 
• Further R&D on smart metering products may benefit from international co-operations and funding possibilities.
• Training, for example, integrating the smart meters installation in regular training of technicians, can address the shortage of skilled manpower for installing the system. Meanwhile, educating users to operate and benefit from the smart meters is also essential (Consumer Focus 2010). 

For more informative and frequent billing:

• It may encourage energy saving behaviour if billing is delivered more frequently and contains the following information (Darby 2006; IFEU 2006): actual meter data, comparisons with previous periods and similar households, appliance-specific breakdown, annual individual energy reports, and advice for energy saving. 
• • Communication with energy companies about their benefits through the measures, for example, that the more frequent billing enables them to better handle non-payment issues, may mitigate their resistance (Duscha & Dünnhoff 2007).

For both feedback measures above:
As part of finding solutions to the cost-sharing among different beneficiaries for the roll-out of smart meters or informative billing, the government needs to provide appropriate regulatory or financial incentives to responsible actors (e.g. network operators, utility companies) to cover their costs.

Recent reviews on feedback programmes in the USA and Canada show electricity savings in the range of 5% to 15%. Studies on programmes in Japan and Denmark indicate lower savings, i.e. 1.5% and 3%, respectively. The wide range of savings may result from different programme approaches and evaluation methods, or to what extent the analyses consider the moderating factors and covariates (Schleich 2011). In addition, different feedback measures in use also result in different savings. Darby (2006) claims that direct feedback (i.e. immediate feedbacks from the meter or display monitors) can lead to more energy saving than indirect feedback (i.e. feedback that has been processed before reaching users, e.g. billing). However, sometimes, a good informative billing programme can result in higher savings than direct feedbacks, given certain socio-economic factors and consumption patterns (Stromback et al. 2011). Furthermore, the potential additional savings from investment induced by feedback measures are still limited. Ehrhardt-Martinez’s (2011) study shows that users mostly decide to adjust their behaviour in response to the feedback rather than investing in energy-efficient technologies.

For smart meters, major costs are its purchasing, installation and operating costs as well as the investment in advanced data collection and data communication infrastructure. KEMA’s recent study gives an overview of the costs of the smart metering system (see table). In terms of the cost for society as a whole, the roll-out of smart metering system will potentially result in an increase of energy price to cover the costs of deploying the technology (Balmert et al. 2012).
In general, the cost depends on a number of country specific factors, for example, the features of the existing local electricity and communication system, customer density, etc. (Balmert et al. 2012; Giordano et al. 2012).
For example, in the UK, the most recent impact assessment conducted by DECC shows, until 2030, the total cost of rolling out smart meters in the domestic sector and small/medium non-domestic sector would be 10,912 million and 592 million GBP, respectively (DECC 2012 a, b).

	Min-Variation	Max-Variation
Procurement costs smart meter (€)	35	110
Installation costs smart meter (electricity) (€)	13	28
Additional installation costs (selective roll-out) (€)	20	40
Annual operating costs smart meter (electricity) (€)	0.4	2.9
Procurement costs MUC-C (€)	60	150
Installation costs MUC-C (€)	8	20
Annual operating/maintenance costs MUC-C (€)	1.7	5.8
Procurement costs smart meter (gas) (€)	70	140
Installation costs smart meter (gas) (€)	25	45
Annual operating costs smart meter (gas) (€)	1.4	3.6
Procurement costs inhouse display (€)	5	50
Installation costs inhouse display (€)	10	25
Installation costs PLC per smart meter (€)	4.5	9
Annual operating costs PLC per smart meter (€)	0.7	2
Additional costs GPRS/GSM per smart meter (€)	35	50
Annual operating costs GPRS/GSM-modem (€)	1	10
Installation costs DSL & additional costs per smart meter (€)	60	100
Annual operating costs incl. queries DSL-Modem (€)	0.4	4

Given their potential for changing users’ behaviour, feedback and other measures targeting behaviour can be cost-effective.
However, for the deployment of new technologies such as smart meters, its cost-effectiveness still needs to be carefully assessed in different countries. In terms of its potential benefits for society as a whole, except for energy saving, it can additionally generate the following benefits:

• reduced meter reading and operation cost
• reduced operation and maintenance cost
• deferred distribution capacity investments
• deferred transmission capacity investments
• deferred generation capacity investments
• reduced electricity technical losses
• electricity cost saving 
• reduction of commercial losses due to electricity theft
• reduced outage times
• reduced CO₂ emissions and reduced fossil fuel usage
• reduction of air pollution

(Giordano et al. 2012)
In addition, smart meters may be crucial for network integration of fluctuating generation from renewable energy sources.
In the UK, the most recent impact assessment conducted by DECC shows that, by 2030, the best estimate for the net benefit of the domestic sector and small/medium non-domestic sector would be around 5 billion GBP and 2 billion GBP, respectively (DECC 2012 a, b). This would mean the installation of smart meters would be cost-effective for the small/medium non-domestic consumers but (on average) not for the domestic sector.