This paper describes the strategy which forms the basis of the organisation of Energydata and cooperation between central and local authorities. The general principles for application and collection of data are also treated. Energydata has successfully been used as a data source when a number of energy projects, analyses and follow-up activities were being planned and implemented. Use is typically made of Energydata by central and local authorities, energy companies, and research institutions.

Before 1970 the main point of departure of Danish energy policy was supply problems in connection with international crises. The then Ministry of Trade had responsibility for these tasks as energy was regarded as a commodity similar to other imported goods. The first oil crisis meant that an energy policy per se was formulated; energy now changed from being a commodity to receiving the status of a political issue. The major issue of energy policy during the 1970s was to ensure a multi-pronged energy supply and to reduce Denmark's dependency on oil. Immediately preceding the first oil crisis, oil represented approximately 90% of total energy consumption.

During the years that followed the international oil market showed a high degree of instability resulting in steep price increases. Rising prices for oil meant that oil imports came to figure largely in the Danish balance of payments. This had a direct influence on the objectives of energy policy; from the end of the 1970s there was increasing emphasis on reducing economic dependence on the international energy markets.

From the end of the 1980s environmental considerations were placed at the centre of Danish energy policy. In Energy 2000 (published in 1990) energy-policy goals were directly linked to the debate concerning sustainable development and international discussions about global environmental problems. In this connection a political objective was adopted to the effect that Danish emissions of CO₂ should be reduced by 20% between 1980 and 2005. The particular focus of this energy plan was the electricity and CHP sector which represents about half of total CO₂ emissions from the energy sector.

In 1996 the Danish Government presented Energy 21, a new action plan for energy. Energy 21 described a number of fresh political initiatives which were to contribute to Denmark living up to the environmental objectives which had been established. Furthermore, targets were established for further limiting CO₂ emissions after 2005. Utilising renewable energy and implementing energy savings were regarded as major areas of focus.

Energydata was launched in the 1980s as a direct consequence of Danish energy and environment policy. Since the 1970s several objectives have been set up for the areas of energy and environment, policy which, inter alia, was formulated in sector planning for heat and power supply, savings initiatives and decisions concerning a number of large-scale energy projects. Implementing the political objectives led to investments in the Danish energy sector amounting to several billion Danish kroner annually.

The high degree of municipal engagement in realising Danish energy and environment policy is due to two main factors: firstly, the sector planning which was established gave the municipalities a number of central tasks and competencies; and secondly, the municipalities were and remain directly involved in the energy supply itself via large-scale ownership of facilities in the district-heating, natural gas and electricity sectors.

This was why it was necessary to ensure that there was a good data source for energy matters both for central planning purposes and when individual projects were being designed in the municipalities. There is, therefore, also a need for geographically-specific information on energy consumption, supply systems and production.

The Heat Supply Act was adopted in order to ensure that Denmark had a multi-pronged, rational supply of energy. This sector plan was strongly influenced by spatial planning with a clear division of labour between central authorities, counties and municipalities. The municipalities were to cooperate with the central authorities on their heat planning. Thus, although a consensus was reached on the data source, there were different objectives of and attitudes to energy policy.

Subsequently, the municipalities have been through a very extensive process of sector planning where areas have been designated for heat supply networks and individual heat supply, respectively. This geographical division has comprised the framework of the very extensive development of the Danish district heating supply. Existing district heating systems in the large towns and cities have been enlarged and a large number of district heating networks have been established in small and medium-sized towns. One consequence of this is that today Denmark has the highest coverage of district heating in the world, 49% Simultaneously heat planning has ensured a market for a nation-wide natural-gas project which was developed up through the 1980s and the beginning of the 1990s.

The municipalities have prioritised forms of supply in the individual areas on the basis of national economic and environmental considerations with the aim of avoiding double supply and thus double investments. This means in turn that the district-heating and natural-gas companies respectively have a monopoly on heat supply networks in the respective supply areas.

The problems of the municipalities in relation to heat planning was the lack of any reliable data about energy matters in the individual supply areas. On the national level the only information available was that which made general planning possible, but this was obviously insufficient in relation to the new planning tasks.

A number of large energy companies had already established or were in the process of establishing their own computer development projects and had a great deal of relevant information about, inter alia, energy consumption. However, the energy companies' initiatives were not mutually coordinated and they defined the data demand in relation to their own tasks. On the other hand the authorities had to take the final decisions with regard to energy policy on the basis of national economic and environmental conditions.

To ensure a solid data source and independence of the energy companies involved, the first municipalities elected to conduct extensive, resource-demanding mapping of current heat demand and fuel use, and to set up forecasts of the future development of the energy demand as a consequence of initiatives on insulation and forecasts for new building. The central authorities and the municipal organisations supported this activity in the form of pilot projects, guidelines and agreements that made it possible to make use of extracts from public registers.

Experience from the work of the first municipalities was not positive; they had used relatively great manpower resources and consultant funds on carrying out the task . On the other hand, the municipalities now had a good data source for carrying out their forthcoming planning tasks which was even better than the information possessed by the energy companies. This was quite important for the subsequent negotiations between the municipalities and the energy companies about the content of energy planning.

Finally, the municipalities were not satisfied with the division of labour between them and the other public-sector authorities. In their view the municipalities carried out the greater share of the work of collecting data for joint use which was placed at the disposal of the counties and the central energy authorities. Furthermore, the data which summed up districts were of limited value when the municipalities subsequently were to draw up detailed projects or to administer planning decisions concerning the individual building/proprietor.

The Danish Energy Agency, in a process of dialogue with the municipalities, drew up a joint strategy for future data collection and cooperation on the basis of the experience of the first municipalities' energy mapping. This strategy has been later developed in step with new tasks and experience gained.

The guidelines below formed the basis of cooperation between the Energy Agency and the municipalities and for the overall computer and database development.

The basic philosophy for further developing Energydata is elaborated in the following.

Cooperation was established between the Danish Energy Agency and the municipalities with the objective of fulfilling the needs of both parties with regard to planning, design and administration in the field of energy. Satisfactory cooperation between the two parties is a prerequisite for data collection to function. The total Energydata concept sets the scene for a division of labour where the central energy authorities and the municipalities join forces to procure, process and utilise geographically specific information about energy matters.

The municipalities' need for data pertaining to energy should be regarded as an integrated whole so that all data necessary for planning purposes are procured. Energydata should be developed in accordance with this. The municipalities have had less favourable experience from other areas of planning where computer-based solutions focused on certain elements such as planning, thus not paying attention to the later phases where more detailed information was required.

Where the Danish Planning Act lays down requirements concerning content and method in planning, the municipalities are free to choose computer-based solutions and data sources, meaning that Energydata is an option for the municipalities. By minimising the work of the municipalities, operating with fixed, low prices when supplying data, and developing the system in step with new municipal tasks, the aim of the Energy Agency has been to offer Energydata to every municipality in Denmark.

The Energy Agency has undertaken to finance, develop and operate an overall computer-based system. To minimise resources, the system is largely based on extracts from public registers and statistical accounts and, to a lesser extent, on reports from the municipalities. The public registers contain information that describes, for instance, what buildings are used for, their size and level of activity. By employing joint identifiers/keys, it is possible to link information about buildings and enterprises from administrative registers to a clear geographical locality in the shape of an address or a coordinate. Information from the registers is used, moreover, as a point of departure when calculating the energy consumption of the building in question.

The registers and calculation of energy consumption serve to minimise the need for reports from the municipalities. However, tasks will continue to exist that presuppose knowledge of local conditions, such as planning decisions and delimitation of geographical areas, which can only be undertaken by the municipalities. The idea of this division of labour is to ensure that the municipalities approve and supplement the material that exists about the municipality. One particular task is to divide the municipality into smaller geographical areas - energy districts - that comprise the smallest units in Energydata.

All in all this means that the work of the municipalities will be reduced to tasks presupposing local knowledge and which do not necessarily require in-depth knowledge of energy-related matters. In return for the reports and supply of data from the municipalities, they can order completed reports and supplies of data from Energydata.

Standardising and simplifying the work of the municipalities will ensure the data source and a greater degree of acceptance on the part of the municipalities. At the same time the central energy authorities will be assured a highly detailed, up-to-date data source which will be possible to use as a basis for overall planning, surveys, research projects and administrative systems.

As the need arises for new kinds of information, or when the municipalities are charged with new tasks, Energydata is obliged to develop new standard products that correspond to these tasks. The need for new types of information has changed constantly. In the course of the 1980s and 1990s energy planning and public regulation in the energy field has been extended to include electricity production, energy savings and industrial processes in trade and industry. Energydata has been developed in step with this from being a data system with primary focus on matters concerning heat supply to being able to deal with energy producers also. The fundamental concept of linking analyses to geographical references and utilising information from a number of official registers and surveys has, however, also proved to be well-suited to handling new themes in the areas of energy and environment.

It has been a basic concern of the Danish energy Agency to underpin planning legislation in the field of energy with a special emphasis on the field of data. As is clear from the above, however, at the same time the central energy authorities ensure that they have an extremely precise data source, approved by the municipalities, that can be used for planning, administration and research. This means that these authorities obtain an excellent point of departure for conducting negotiations with the municipalities. Thus both local and central authorities obtain a uniform data source for taking decisions as there is agreement concerning the description of energy consumption and supply.

The work of designing a computer-based system with energy information opened with a clarification phase concerning the way in which information concerning energy consumption, supply and production should be dealt with. Defining relevant information and contexts was a central phase in the development of Energydata where traditional database design is combined with GIS theory.

A database/computerised model that reflects a physical system must of necessity be based on a number of simplifications in relation to reality. It is neither possible nor desirable for a database system or an analytical model to describe all the features and characteristics that exist. For this reason the objective of the design phase is to identify the parts of the real system that should be included in the model and the parts that are to be excluded. This delimitation of content is essential for the following development of the computerised system and for the application and flexibility of the model. Less positive experience of analytical models can often be ascribed to the tools being used outside of the intended areas of application.

There is no absolute yardstick for measuring what a good database design is. An appropriate model for the Danish energy or environment system is an information system that contains the information asked for by its users. However, needs change over time. As Energydata supplies data to a wide range of customers for a wide range of tasks, it is especially important that the system can be developed in step with the tasks and the information demands of the users.

Energydata has been constructed in line with the "bottom-up" philosophy. This means that all information is registered on the level of the smallest unit; it can then be summed up on an arbitrarily aggregate level, for example on an administrative division like a municipality or the catchment area for an energy company.

The immediate advantage of bottom-up systems is that the data are consistent at both detailed and general levels. Users have ample opportunity to set up their own search criteria and delimitations. The disadvantage is that it can be difficult and require many resources to compare and maintain the great amounts of data involved. At the same time there are high quality requirements regarding the information entered into the system, as the summed-up data should be in accordance with the basic data and coordinated with corresponding national figures.

Contents of Energydata

The requirements of Energydata are described in further detail below.

The objective of Energydata is to secure precise data about energy consumption divided into:

• areas of consumption (space heating, electricity consumption, industrial processes)
• categories of customers (housing, public administration, service, industry, supply companies)
• types of energy (fuels: oil, natural gas etc., and production of secondary energy: electricity and district heating).

Energydata is also to contain information concerning energy production for the individual production and distribution plant. This includes information about form, volume and capacity of production, and consumption of primary fuels. Finally, the supply areas of the large energy companies are registered in the database.

Energy statistics can provide a picture of current conditions. However, this type of information cannot be used for tasks involving planning and design and for administrative purposes without supplementary knowledge of planning decisions and other measures that influence the future development of energy consumption and production. This may consist of decisions to reestablish or convert an energy production plant, prohibiting certain energy technologies, or requirements concerning the efficiency and use of fuel in new energy plants. Linked to information about consumption and supply is information that enables accounts to be drawn up directly targeted at the information demand of the local authorities.

It is a precondition for Energydata being used for analyses of energy matters in specific geographical areas and by local authorities that information with regard to consumption and supply has a place-bound reference.

Energydata is an information system that is to receive and supply information in relation to a number of external partners. To facilitate pooling with public registers or the client registers of the energy companies, the individual objects must be provided with the common identifications or keys. Similarly, the geographical references must be close to the usual standards for GIS.

Energydata must contain validated and mutually coordinated information about energy consumption and supply. The result of this coordination is an annual edition describing total energy turnover for the year in question.

Energydata is a GIS system where all information refers to geographical locations - areas (polygons) or points (nodes): for this purpose Denmark has been divided up into over 5,000 energy districts. The basic elements of Energydata are described in the following.

In Energydata buildings and technical installations carry information as estimated and/or actual information about energy consumption and production are linked to them. Buildings and technical installations comprise the smallest nodes of information. The point of departure is that energy consumption and production take place in or are directly connected to buildings and/or technical installations.

A technical installation is defined as a permanent technical construction with energy consumption and/or production. Each year the Danish Energy Agency collects information about the technical installations that produce electricity and/or heat (cf. the section below on energy producers).

The database contains information about all 2.4 million buildings in Denmark. Once a year the information is transferred from the Building and Housing Register (BBR), a joint municipal and state register of all buildings in Denmark. The database contains information about the structure and technical installations of the buildings. There are, moreover, data about primary use and type of ownership. The buildings have a geographical locality in Energydata as the individual buildings refer to the appropriate energy district.

Most GIS database systems have a one-to-one relationship between database objects and the digital maps. Buildings/technical installations are the smallest point unit in Energydata database, and energy districts are the smallest area units, corresponding to a one-to-many relation where on average there are 500 buildings per energy district This relation is quite intentional as it has been decided to demand more of the database's description of consumption and production than of the geographical location.

In practice it is unrealistic to collect information about the actual energy consumption of millions of buildings. In a number of cases, on the other hand, it is possible to estimate expected annual consumption reasonably accurately on the basis of BBR information about the size, structure, heating installations and use of the buildings . This applies to the household, service and public sectors in particular. Reference is made to the section on large-scale consumers for production enterprises.

Specific unit figures for the heating sector and for a great number of types of buildings and uses have been drawn up for purposes of calculating expected energy consumption. The Energy Agency commissioned the Danish Technological Institute to construct a heat consumption model that establishes specific unit consumption figures (GJ/m²) for the heat demand in buildings on the basis of BBR information about type of building, area and storeys, building materials and date of construction.

The unit figures used have been coordinated at the national level on the basis of statistics about total energy consumption and the associated areas for the respective sectors.

It is not meaningful to estimate expected energy consumption for large business enterprises, power plants etc. on the basis of information about the size and character of the buildings in question. In addition, these buildings may consume more energy as energy is used for industrial process and/or energy production. For this reason, in the case of this type of energy consumer information is collected about the actual consumption of the enterprises in question. These are termed large-scale consumers in Energydata and include buildings with big or atypical patterns of consumption. In this way, apart from more precise information concerning energy consumption, information is procured about type of fuel, the type/branch of enterprise, capacity etc.

Energydata contains information about a total of 15,000 large-scale consumers. The information, which is gathered from public registers and own counts, is up-dated once a year. The large-scale consumers are registered in Energydata as separate objects and with a reference to the associated building.

Producers of electricity and heat comprise a special type of large-scale consumer. In the case of these plants, information is needed about information concerning energy production apart from information about actual fuel consumption. This has mainly to do with information about production technology, capacity, production of electricity and district heating, and the network to which the energy produced is supplied.

The design of electricity and district-heating production is decisive for energy consumption and environmental impact in Denmark; electricity production is responsible for approximately 50% of total CO₂ emission in Denmark. As it has not, however, been possible to retrieve this type of data from existing registers, once a year the Danish Energy Agency conducts a mapping of electricity and district-heating production where all sites of production and plants are registered. In 1997 the account included over 1,100 production sites with a total of 1,600 electricity and/or district-heating producing plants (excluding wind turbines).

In the case of energy producers, information about energy production and supply, the relevant district-heating network, matters to do with ownership and geographical location are registered in connection with a coordinate (point).

Figure 1: Relations between energy company, plant, installation and district-heating network

Conceptually, the energy producers have been divided into plants and installations. Links to other objects are built up around the concepts of company, plant, installation and district-heating network (cf. figure 1).

An installation is the machine where energy is converted to electricity and heat. All information about energy production is attached to one, and only one, installation. With a starting point in the type of installation, the installation's energy/fuel input and electricity and/or heat output are validated.

A plant is the term employed for a site of production at whose address one or more energy-producing installations are to be found. Different plant types have been defined. Furthermore, a plant is geographically identified with a property and building number in relation to the BBR.

The relation between energy producer and energy district is an example of how overall point and area objects can be combined in Energydata. An energy producer with a known position (point), supplies district heating to an energy district with district heating supply (district heating network). The network, which is an area concept, is supplied by one or more energy producers.

In the light of the key importance of the energy producers for energy supply and environmental impact: the geographical localisation is, in addition, coupled to energy districts identified by a coordinate for the location of the energy installation (point).

Energy districts comprise the smallest area unit in Energydata. An energy district is an area with a uniform structure of energy consumption and supply. This may be, for example, a housing area supplied with district heating, an industrial area with the possibility of natural-gas supply, or a farming area with scattered buildings. The individual municipalities have undertaken the geographical division. The typical municipality is divided into 15-20 energy districts.

Smaller towns usually comprise separate energy districts whereas the larger urban areas most often consist of several districts, reflecting different types of buildings and/or areas of application (housing areas with one-family houses, city centres with blocks of flats and service industries, industrial areas etc.) and different types of heat supply networks. Other geographical objects such as towns, municipalities or the heat supply network are defined by summing up one or more energy districts.

Each energy district is provided with a clear identification (key). Information is registered about planning decisions and forecasts for new building and future energy supply.

Since the 1970s Denmark has had a register of all roads and addresses - the Central Personal Register (CPR) which contains information about all 12,000 roads in Denmark and their approximately 5 million addresses. In establishing Energydata the CPR was supplemented by references to the energy districts. The enlarged CPR makes possible automatic geographical coding in relation to existing divisions into energy districts. This enlargement of the register, undertaken by the individual municipalities, means that more than 99% of all buildings in Denmark can be clearly placed on the energy district map.

In addition to a register dimension, Energydata also has a geographical dimension for the smallest units. At the moment energy districts and energy producers are geographically localised. In future addresses in the BBR will be localised meaning that the other objects in Energydata will also have a geographical dimension. There are thus two possible ways of linking information concerning energy consumption and production to the energy districts: either by stating the official address of the buildings in question, or by stating the geographical coordinates (UTM coordination system).

All district heating networks in Denmark are registered in Energydata, inter alia with a view to analyses of CHP production and sales of district heating. Included is a total of 380 technically separated networks supplied by approximately 600 CHP and district heating producers.

A district heating network is a physically (technically) connected district-heating distribution network that receives heat from one or more heat producing energy installations. The supply area of a district heating network is defined as one or more energy districts in the municipalities where a distribution system has been established. A district heating network offers to supply heat to all consumers within the energy districts of the network. All buildings in Denmark are related to an energy district, it is possible to calculate the total heat demand and types of buildings in each district heating network.

As all consumers in Energydata are included in an energy district, consumers can be divided by location in relation to the individual district heating network. This means that it is possible to identify the buildings that receive heat from specific place of production and installations and, similarly, that it can be established which buildings that can immediately be connected up to existing distribution networks.

The basic elements of Energydata are directly linked to physical production of energy. In a number of contexts, however, it is relevant to describe energy matters and associated environmental effects summed up in administrative divisions. This may, for example, be an account of fuel consumption in a given town or city, or total CO₂ emissions in a municipality. It may be interesting in other contexts to analyse the distribution of energy consumption on forms of energy in areas where decisions have been made ordering or prohibiting specific types of heating.

For these purposes Energydata contains a geographical representation of the different administrative divisions which are represented as divisions of area, such as towns, municipalities, counties and provinces.

Both the energy companies, that is district heating, electricity and natural-gas companies, and the local and central authorities make use of information from Energydata.

An energy company is the legal entity that owns one or more energy producing installations. In addition to this, an energy company may have a number of functions in relation to the energy supply, such as production, transmission and distribution of electricity and/or heat. An energy company may also be a production enterprise that owns an installation and produces electricity and heat alongside its primary production.

Supply areas for different distribution and transmission companies are registered in Energydata making it possible to conduct analyses of the composition of the markets of the individual energy companies. As is the case with the administrative divisions, the supply areas of the energy companies are defined on the basis of Energydata's energy districts and for this reason as expressed as areas. The ownership of the energy-producing installations is described in order to supplement the registration of the supply catchment areas.

The above descriptions deal with the basic elements of Energydata. Figure 2 shows the logical model for Energydata; the key objects are emphasised.

The points of departure in this bottom-up model are, respectively, the smallest point unit (buildings/technical installations) and the smallest area unit - the energy districts. Conceptually, the objects may be divided into a number of thematic units linked together by keys.

Figure 6.4, an overview of Energydata's objects, consists of the following areas

A number of validation rules have been set up for the content of the logical model where the intention is to secure a consistent data model. The model is built up in several layers where the general concepts are generated on the basis of the fundamental elements described -energy districts and buildings.

It is an invariable rule that all the key relations must be present. This means, for example, that the logical model is unable to deal with large-scale consumers or energy producers that lack relations to buildings/technical installations. Similarly, it is a precondition for area units that they have a relation to at least one energy district. Another example: a district-heating network cannot exist in the model without a production installation connected to it, or district heating producers with no information about the associated district-heating network.

The above-mentioned validation rules imply that the quality of the information used must be high. The demand concerning the data contained here is a consequence of the bottom-up strategy that has been adopted. The moment that all information is linked to the smallest units reported and when users get the opportunity to establish search criteria and area delimitations, any mistakes in the data become extremely visible. In brief, the logical model and validation rules are inextricably interconnected.

In view of the large volume of data, the extensive validation and the many actors involved, including all the municipalities that contribute to the up-dating, energy information as a whole is organised as an annual cycle, the individual phases of which are described in the next section.

The result of the annual cycle is a statistical description of the coherent energy system which is known as the annual edition of Energydata. This description of the connection between energy consumption and supply is the final result, forming the basis of energy analyses, planning tasks, and administration during the following year.

There are some clear advantages of the form of cooperation that has been chosen. Firstly, it ensures that all actors in the field of energy work on the basis of identical data. Secondly, data can be stored in a form well-suited to analyses and integration in the users' own information systems. Thirdly, validation can be completed before the energy information is communicated further.

It has traditionally been regarded as good database administration to store information so that it takes up the least possible space with a stated degree of detail. This means that an effort is made to store the information in one place only thus avoiding double registration (redundancy), primarily with a view to maintenance of the database. This is unimportant in the case of an annual edition that does not have to be maintained subsequently. On the other hand, there may be a need to compound information according to certain tasks when data is subsequently processed which may lead to the same information being present in different data extracts. The main consideration in this situation is now to ensure that the information that has been compounded is well-suited to analyses and further processing. This allows users easy access to relevant information and they can select precisely the topics (sets of data) that are relevant. In some connections this is referred to as a data warehouse.

The information in the Energydata annual edition is registered in accordance with the above principles where there is a stringent maintenance section and a redundant annual edition. The former is constantly up-dated in connection with the annual counts and entering of data from public registers. The annual edition is produced when the data in the maintenance section have been checked.

It is important to keep the maintenance section separate from the annual edition. In addition, a stringent logical model should be established to develop the computer-based system and, as a basis, be laid down by the logical validation rules.

The philosophy and concept that formed the starting point for Energydata's strategy and development have been described in this paper. As can be seen, creating well-functioning cooperation between central and local authorities and other actors in the energy field has ben emphasised. In reality, the strategy is proactive and requires relatively many resources as the central energy authorities have undertaken a number of functions that normally lie outside the scope of governmental administration.

There are a number of examples of computer-based projects where the information demand has been established in advance, making the development more straightforward. On the other hand, the general disadvantage of such systems is that they quickly become obsolete when tasks and the data require change. When Energydata was being developed, the initial phases focused on the form and obligations of cooperation, the potential for making use of public registers and keys, and a geographical approach. With a basis in these agreements, it has subsequently proved possible to develop the computer-based system and cooperation in stages, including integration with digital maps. This means that today the system is a Geographical Information System (GIS system).

The basic line of thought is that the database/GIS system should be based on objects and logical rules that are relevant in both purely energy-technical physical matters and in planning and organisational terms.

In an open system of information there are great demands on the logical model: it should be possible to develop it in step with new planning tasks and associated information needs, at the same time as existing data sources set limits to the information that can be procured.

It should be emphasised that a seemingly complex model with extensive validation rules does not necessarily mean that it is just as complicated to use the information system. The Energydata system contains a number of overall concepts (towns, municipalities, energy companies). These area concepts have been introduced to allow users to choose between familiar objects when requesting information about energy consumption and production. In continuation of this, the philosophy behind the data warehouse is that the information should be communicated further in a way that is appropriate and possibly redundant for the users.