• contribute to the improvement of overall efficiency of energy use;
• contribute to a better use of energy resources and assets available at the electricity supply side;
• contribute to the EU environmental goals, mainly through demand-side management in the context of urban environments.

Hence, sustainability may be pointed out as the general framework of this project. It has been directed to urban realities, and cities are one of the most important targets of the EU energy and environmental policies. Energy market transformation is happening through a number of mechanisms and it is very important to influence the course of events towards sustainable development. Although a multitude of aspects may be considered in this context, demand-side intervention is certainly one essential component of the above mentioned policies.

The project dealt with this type of issues, in particular with utility-driven demand management, which is facing a whole set of problems that did not exist prior to energy market liberalisation. Utilities have presently different stimuli of those that existed in the past, in a business environment where the concept of natural monopoly could sometimes be beneficial to demand-side actuation, through revenue reconciliation procedures or simply through government mandates when companies were publicly owned. Nowadays, market transformation informed by energy efficiency-driven policies, in a freer market framework, is broader than demand-side management as this one has been defined in the eighties, but nevertheless includes many of its mechanisms.

The partners involved in the project have been

• INESC - Pólo de Coimbra
• Cenel - Electricidade do Centro
• Universidad Politécnica de Valencia
• Dep. of Electrical Engineering, universitá di Roma "La Sapienza"
• Fraunhofer-Institute of Information and Data Processing (IITB-EPS) - Dresden
• Universidad de Murcia

The project has been developed centered in two different cities: Coimbra, in Portugal, and Dresden, in Germany.

In the first case, the motivation has been the existence of an urban substation of the medium voltage network that the utility fears will have capacity shortage in the mid-term, as there has been a recent Winter where maximum demand reached an extreme value superior to 80% of capacity. Hence, possible demand-side initiatives should be evaluated in order to balance the two options: supply side capacity expansion and demand-side management (DSM).

In the case of Dresden, the launching of a utility's programme towards telemetering has allowed an extensive data collection with a fine time resolution of power demand values, as well as a pilot-experiment with direct load control.

These two cases were, hence good opportunities to assess several ways of acting at the demand-side, involving a whole set of problems and procedures.

From a utility's point of view, capacity shortage at the supply side is still an opportunity to consider demand-side resources as alternatives to supply-side capacity expansion, especially when strong financial constraints exist. There are, however, important problems to face when considering these alternatives, namely related to the choice of the possible measures to be implemented, depending on its potential effects on demand level, and to the results of cost-benefit assessment. In both fundamental aspects it is essential to know relatively well the structure of demand and certain other variables that allow the characterisation of consumer attitudes towards energy services, utility's initiatives, technology options, etc. It is also important to develop means of estimating or predicting effects of certain measures without needing to perform market inquiries and load research every now and then, as these are expensive and time-consuming activities. Load models and demand models are appropriate tools in this matter. Indispensable is also to have load research data from campaigns designed to provide appropriate information on load behaviour and typical demand patterns, in order to build an information background to the studies and reasoning referred above. The project has given the opportunity to explore this important side of market transformation which is dependent of utilities' initiatives.

On the other hand, still from the utilities' point of view there are also important opportunities to demand management through the availability of telecommunication facilities. In this case, demand management is one of a set of services to the consumer that communications are able to provide. However, strictly from the point of view of demand management, two-way communications offer a great potential, not only for consumption monitoring through remote meter reading, but also for remote control applications that allow an effective capability of managing demand and improving distribution network operating conditions - leading to a more rational (sustainable) use of supply resources. The potential application of these technologies depend on the degree of development of telecommunications in a given area and on the evaluation that a utility makes of the business opportunity of investing on such facilities. Power-line carrier based systems, for example, are being more and more applied for these purposes and have the significant advantage of not needing a new dedicated medium for data transmission.

However, the global perspective of urban intervention has also been present. An action guide has been developed for elaborating and implementing energy plans in cities, according to best practices known world-wide, in order to provide a self-contained package to be applied wherever local authorities engage in sustainability programmes that include energy efficiency measures.

Various lessons have been learned throughout the course of development.

Firstly, methodologies for load research have been perfectioned, as media specifically conceived for data collection have been extensively tested and experience has been accumulated on its use. Contacts with customers provided an invaluable experience on how to address consumers and on their expectancies, knowledge gaps, common sense, etc. Load research provided, of course, its main product, which is a deeper knowledge of demand structure and of its dynamic behaviour. Exploration of the collected data is now a wide perspective, as data processing may be conducted still on a number of different directions that have potential usefulness to detect cross influences among variables and thus lead to perfectioning the definition of some demand management strategies.

Secondly, load modelling has shown it is possible with a moderate amount of permanent metering points in the network to estimate load patterns virtually in any point of interest in the network - opening the possibility of estimating load diagrams at various demand aggregation levels. It also has shown that with bottom-up approaches based on psychological behavioural models it is possible to tune simulators to tightly reproduce demand behaviour - at the levels of load diagrams and of energy consumption - of several aggregation levels and categories (consumer classes, end-use equipment). These models are also very effective in anticipating the impact of demand management strategies, providing a reliable way of anticipating results of DSM actions before actual implementation. This is common to physically-based load models, which have also been used to evaluate the potential of influencing demand at aggregate level through end-use load management (as is the case of electric water heating loads). An example of the accuracy of such models can be seen in the following figures. The left one shows the average dwelling load diagram of the target area as obtained from actual data collection. The one on the right is an example of the simulated anticipation of the effect of lighting retrofit for efficiency improvement for various penetration levels, where it is also possible to notice the match between the upper curve (simulated) and the one in the left figure.

Thirdly, for the first time in the target region of the pilot case of Coimbra, DSM has been implemented through direct installation of energy efficient equipment by the utility in consumers premises, providing valuable information not only about the actual effects of the replacements but also on the problems and difficulties associated to this implementation method (due to barriers either on the consumer side or on the utility side) and to the respective monitoring of results. As regards to the physical impact of replacements, they showed to be potentially very effective in view of the utility's load shape objectives at the target substation level (demand limiting and valley filling/load shifting), mainly through energy efficiency improvement.

Also, a pre-screening method of DSM actions has been applied, providing results that, on one hand confirmed previous reasoning based on load research results and, on the other hand, allowed a systematic view of the interactions among the various variables that influence DSM evaluation. It showed to be a useful tool for preliminary definition of DSM programmes, through the selection of the most promising end-use technologies and implementation methods.

The practical implementation of telemetering and control of end-use equipment has demonstrated with extensive data collection, in Dresden, the advantages of the use of telecommunication facilities for demand management. This experiment also allowed developing a load model suitable for predicting the effects of thermostatically controlled loads on power demand, which is a well-known effect when using remote load control.

The project dealt with several aspects of demand management, as load research, load modelling and simulation, DSM impact on demand, DSM programme evaluation.

Load research has shown to be indispensable to assess possible demand-side interventions. It has been conducted targeting both physical and behavioural aspects of demand, which has proven to be very important if subsequent load modelling is to be used to anticipate impacts of demand management actions.

Also, the pilot experience on actual DSM implementation showed that adequate monitoring before and after intervention provides the possibility of identifying potential impacts of some measures. In the particular case of Coimbra, it showed some evident potential influence of lighting energy efficiency improvement in peak demand at the target substation level.

Load modelling methods used in the course of the project must be retained for demand management activities planning, as they showed that a combination of approaches answers a set of different questions that must be addressed. Namely, taking the most of load research and monitoring has shown to demand a careful planning both of metering points to be instrumented in the electricity distribution network and of field activities for data collection with consumers. Load models may help in:

• the identification of average load profiles of consumer classes;
• the identification of average load profiles of different end-use equipment;
• the possibility of knowing the load profiles at key points in the distribution network;
• the anticipation of effects of some demand management actions prior to actual implementation or experimentation, thus avoiding high evaluation costs;
• when using remote load control procedures, the definition of control strategies that avoid problems associated to the payback effect of loads with local thermostatic control.

There is a good potential for utility-driven demand management. However, it must be based on sound evaluation of costs and benefits, for which the capablity of anticipating barriers and effects of demand-side actions is indispensable. The project demonstrated the use of such tools. However, for the sake of suatainability, local authorities must get involved in the global effort of market transformation. The action guide for urban energy planning presented in the project's report may be of help.