C I R E D17th International Conference on Electricity Distribution Barcelona, 12-15 May 2003
Special Report – Technical Theme 4
Distributed Generation - Management and Utilisation of Electricity
Tony Headley
Chairman – United Kindgom
Andrew Cross
Rapporteur - United Kingdom
Andrew CrossSession 4Special Report- 1 -
C I R E D17th International Conference on Electricity Distribution Barcelona, 12-15 May 2003
Continuing change in the utility and energy markets combined with public support for environmental issues has insured that the topics covered in session 4 of CIRED continue to create strong interest. Significant numbers of generators of all types have been connected within distribution networks encouraged by legislation opening up access and giving incentives to renewables and lower CO2 emitting equipment. This growth shows no sign of decreasing. There is high international interest in the experiences of the generator developers and network operators. These drivers not only impact upon generation and network operation but also upon planning, load prediction and demand management. Liberalisation of the power market and restructuring of utility organisations are changing the way in which end use energy efficiency goals are being approached. Examples of how these challenges are being met complete the scope of this session.
With the larger number of papers in the 2003, the session format has extended as follows:
- The main Alpha day (Thursday 15th May) will first cover the issues in managing and using electricity. Then blocks 2, 3 & 4, will cover the greater number of papers and topics in dispersed generation. A keynote presentation will introduce topical issues at the start of the block.
- On the Beta day (Tuesday 13th May) there is the opportunity to attend the Distributed Generation Round Table where the authors of papers covering modelling will discuss their approaches with a panel containing leads researchers in the area. This will be followed by presentation of leading work from the European Union research programme.
- There is also the opportunity to discuss with authors on a one to one basis and see demonstrations of their work in the Interactive Forum on Wednesday 13th May.
Alpha Day blocks start with a keynote presentation, then authors (marked with *) will present the updated highlights of their work before all the authors form a panel to interact with the delegates in the discussion period. All delegates are encouraged to prepare in advance contributions to the discussion and answers to the questions below.
Block 4.1: ENERGY EFFICIENCY, STORAGE, DEMAND & METERING
The keynote presentation for the first block is Paper 4.10*. This describes how systems can be implemented for demand management and demand side bidding in the Norwegian deregulated and market based power industry. In recent CIRED conferences, these issues have been mentioned in many studies examining the future, this paper marks a step forward in describing systems in use and delivering real commercial benefits.
End use energy efficiency
Paper4.1 describes an electrochemical process to treat the leachate water from landfill sites. Performance of the laboratory scale system is compared. The prototype industrial scale trial will show if this process will reduce pollution problems on a larger scale.
Papers4.3, 4.4 and 4.5 describe three different initiatives in the Egyptian energy sector with aims of both improving energy efficiency and reducing emissions. The first (4.3) covering both load management and end use programmes whilst the second paper (4.4) shows how energy may be saved in natural gas networks now increasingly used to supply dispersed generation. Paper 4.5 examines the environmental and economic benefit in recovering ‘waste’ heat from diesel generation for use in chilling plant, the so called Tri-Generation approach.
Paper 4.6 from Spain describes an attempt to increase the energy efficiency of an existing old building by installation of lower energy lighting systems. The importance of a good knowledge of usage pattern and costing the alternatives to find those with payback in a small number of years is clear.
Energy efficiency in the network
Whilst energy saving in the end use of electricity is described in the papers above, Paper 4.7 shows how the deployment of energy efficient transformers within the Polish distribution network could save 6GWh of energy if all transformers were replaced. The additional cost of the transformers is significant (50% at present). The trading of emission reductions (tonnes CO2 ) can assist in the viability of the programme.
Question 1:
What examples are there of legislators and/or regulators giving strong enough signals for investment in network energy efficiency to achieve real improvements?
Storage Systems
Developments in electrical energy storage systems and the new drivers from competitive power markets have lead to a range of potential applications in distribution networks. Storage systems have been claimed to improve overall energy efficiency by meeting peak demand, reducing the amount of spinning reserve required and assisting in the integration of renewables.
Paper 4.8* describes trials of two types of storage systems one a redox flow battery based energy storage system for peak lopping capable of 42kW for 2 hours. The second a high speed flywheel storage system for power quality (100kW for 100 secs). These studies at CESI in Italy give a valuable insight into the issues of integrating these types of systems into a network.
Question 2:
Availability, capital cost and operating methods will be vital for storage systems to be accepted. Can the authors of 4.8 comment upon these for the systems they have studied.
Demand Side Management
Electricity utilities world-wide are faced with conflicting business drivers as restructuring and open power markets are introduced at the same time as public opinion and governmental action looks for improved energy efficiency to meet Kyoto CO2 goals. The papers in this section along with the keynote show how these challenges are being tackled in different countries.
Paper 4.11 from Italy studies the way to achieve an optimal load management strategy in a pool based market situation with choices on co-generation and storage as well as conventional load management.
Controlling domestic loads (HVAC, water heaters or energy thermal storage heaters) has been implemented in many countries. Deregulated markets can alter the drivers and benefits, paper 4.12 studies the potential in Spanish networks. Given the results, what would the author’s route to implementation be?
Paper 4.9 gives demand forecasting algorithms for weekly and daily demand related to temperature. Examples are shown for a large town in Romania.
In paper 4.13* the issues of DSM in forward and spot markets in Germany is described. It shows how the drivers change from technical to commercial and postulates that in mature markets DSM can be used to increase profit and control risks.
Question 3:
Can the authors and delegates comment upon how the cost of implementation and the risk of changes to regulations or market rules would affect DSM strategy?
To implement liberalised power markets where customers have choice of supplier requires significantly greater functions from metering systems, for example hourly or half hourly consumption figures for many more locations. Papers 4.14* and 4.15* describe solutions to these challenges. In the first, a retrofit solution for Ferraris type meters with communication over GSM is shown to be efficient and capable of collecting data from 7000 meters in part of Denmark in half an hour. In the second paper, an open standard for collecting metering and other data from customers over the web is outlined. This development in an EU research project could significantly improve the ability to carry out these tasks in an open market environment. Experiences in Norway look positive.
Question 4:
In implementing meter data collecting systems for large numbers of customers data quality, system performance and security are key issues, can the authors or delegates give advice for others to follow?
Block 4.2: DISTRIBUTED GENERATION – STANDARDS AND GENERATOR ISSUES
The integration of distributed generation into distribution networks has been new topic that has gained an exposure at recent CIRED conferences. This year we can see a maturing of technical standards especially those covering basic connection issues. We now see documents being used for all sizes of generator from the latest large wind farms to domestic size CHP or PV panels in many countries. This section gives a good overview of these activities.
The keynote presentation for the second block is Paper 4.18*. The CEN Workshop Agreement is the method through which CIRED has contributed to the specification of the electrical interface for <16A per phase domestic cogeneration systems.
In paper 4.16 we hear of the hurdles that were overcome in the USA in creating the IEEE P-1547 standard for interconnecting distributed resources (generation and storage systems) with the power system. This was obviously a large undertaking as agreement was needed on technical issues from a wide range of interested parties!
A different approach to producing national standards for connection of generators of <16A per phase is given in paper 4.19* from the UK. Perhaps by setting a rather smaller scope than that for IEEE P-1547, the joint working group was able to publish a standard for use in a shorter timeframe.
The issues from a Brazilian perspective are covered in paper 4.17. An overview of the regulatory requirements to meet technical conditions for generator connection in France is given in paper 4.21.
Question 5:
Are the regulations and connection standards consistent across countries and are they yet leading to reduced costs and economies of scale in equipment production?
As the amount of generation connected within distribution networks increases the challenges for network operators grow. These have been seen particularly in regions with large numbers of wind turbines. First paper 4.22* examines the challenges for a UK network operator from technical, commercial and regulatory perspectives. Then the following papers cover wind farm specific issues. Paper 4.25* from Belgium reviews the demands on transformers installed in wind turbines and some of experiences where there performance has fallen short of what is needed. A new solution to power quality issues on large wind farms is described in paper 4.26. This use of a synchronous generator with a differential gear is an interesting approach. The reliable performance of installed systems over a number of years will be vital to it gaining acceptance and commercial success. The power quality impact of two different types of wind generator (double fed induction generator with gear, inverter coupled synchronous generator with/without gear) is investigated and measured for installations in Germany in paper 4.28.
Block 4.3 DISTRIBUTED GENERATION – OPERATION
In this block we move on from connection standards and regulations to look at issues affecting the operation of both generators and the networks to which they are connected. Paper 4.32* describes how industry bodies are approaching three key issues in the UK: Fault levels; Voltage control and Power flow management. The focus being on appropriate and cost effective solutions that can be introduced in the short term. There is also a need for operational procedures that are agreed by power producers and the network operator, Paper 4.36 describes those recently introduced in Brazil.
Continuing with voltage control Paper 4.31 examines the interaction of Distributed Generation with voltage regulators from the USA. Paper 4.33 describes a new design for 11kV voltage controller using a statistical state estimation algorithm and control method that can allow increase in generator capacity and double the energy exported. Another approach combining automatic voltage and power factor control is analysed in Paper 4.44 and compared with line rise compensation. The simulations show an improvement in voltage profiles for little additional expenditure.
The increasing size and number of wind farms give rise to voltage and stability issues. Paper 4.35* describes simulation and experiments on windfarms and networks in Spain concluding that double fed Induction Generators are to be recommended.
Question 6:
Are there now cost effective solutions for all the voltage control challenges of renewable generators in distribution networks?
Small scale hydro power is seen as a very green power source, Paper 4.37 examines fault current protection for a valley with some 50 generators totalling 2MW in Slovenia. It shows that the impact is limited and has been successfully managed by the distribution company. Simulations of increasing amounts of DG on both rural and urban LV networks in France are described in Paper 4.38. Some recent EU research project results looking at power quality effects of inverter connected PV generators in described in Paper 4.39, concluding that flicker should not be an issue. An overview of operation, control and stability issues is given in Paper 4.40 which can form a good introduction to these issues for a newcomer to the field.
Fault passage indicators (FPI) are widely used on MV networks. Paper 4.42* examines the impact of DG on their operation, concluding that directional FPIs can be a solution for the moment.
Integrating DG into network management and control systems is going to be an increasing requirement. To assist this process communication and information structure standards are vital. Paper 4.41 describes their development and application to a wind farm demonstrator in Spain. Going a stage further to the integration of DG and storage with Energy Management Systems we see in Paper 4.43* successful algorithms for forecasting both customer load and renewable generation some 24 hours ahead. Enhancing the control capability of the network as an approach to enabling larger amounts of DG to be connected is often referred to as Active Management (AM). Paper 4.45* from the UK presents an outline of some means of achieving this goal. Two further Papers 4.46 and 4.47, from the same team look at protection schemes and a testing facility for AVC relays towards this goal.
Carrier signal based protection to prevent islanding is an approach favoured in some countries whilst regulations in others do not allow it. The use of Distribution Line Carrier (DLC) to achieve this on MV networks in Italy is analysed in Paper 4.48*. Measurements have verified the calculations that distances of 10-15 kms are possible.
Paper 4.49 proposes control algorithms for grid connected inverters and shows simulations of the performance of two systems under fault conditions.
Transient phase to phase faults (e.g. wind blown debris) are very common and reclosing is widely used to minimise loss of supply. An experimental study of how DG on a feeder would affect whether the arc extinguished within the set time is reported in Paper 4.50 from Argentina. The conclusions are positive for those range of faults applied.
Question 7:
How can the variety of individual techniques described in papers above be combined into an implementable toolbox for network operators facing the challenges of running systems over which they have less direct control?
Block 4.4 DISTRIBUTED GENERATION – PLANNING
The first paper, 4.51, in this block proposes a cell based concept for designing networks that can easily accommodate distributed generation.
This is followed in paper 4.52 by a study from France of impact of wind farms on island power systems and how a significant proportion of wind generation capacity can be best accommodated. Paper 4.53* takes the issue of small networks a stage farther describing how low cost microgrids are being designed and constructed in China with the assistance of EdF. An approach to optimising the control of small networks in developing countries with some generation and battery storage is given in paper 4.54.
Monte Carlo simulation of LV connected generation is used in paper 4.55 to study voltage limits. These cover a variety of seasons, loads and generation to estimate the chance of exceeding the voltage limits.
We now have a set of papers oriented to planning in networks with wind generation. Voltage control for networks in USA with wind generators is simulated in paper 4.56. In paper 4.59 the potential of integrating storage and using MV DC technology in networks with wind generation is examined. A number of benefits are identified in reliability, power quality and avoidance of reinforcement. However much more work is needed to create a dispatchable virtual power plant that is cost effective. Meshed operation of MV feeders with wind generation is proposed in paper 4.60* from Finland, quoting significant increases in the power exported above the normal radial operation. The issues being addressed in the expansion of wind generation in Austria from large wind parks are described in paper 4.61. Both planning requirements and future trends are examined. With the large amount of hydro power in Austria there will be opportunities for it’s use to offset the variability of wind.
Paper 4.62 looks at the impact of domestic scale CHP (DCHP) on a UK distribution network. The conclusion being that technical challenges are manageable but the commercial and regulatory issues could be significant.
Question 8:
Forecasts of large numbers of domestic CHP installations by the year 2010 were made at CIRED 2001. Would the authors of 4.62, 4.65, 4.18, 4.19 and the delegates like to give their views two years on?
There were a number of papers at CIRED 2001 describing the needs for better planning tools for accommodating distributed generation and some prototypes were described. This conference we have one paper describing an add-on to an industry standard power flow package, paper 4.63*. It is claimed to give great assistance in planning for the maximum generation capacity. A different approach developed in Brazil is described in paper 4.66* of a tool specifically developed for this purpose.
An all encompassing model of generation plant covering costs, availability, regulatory and market issues in Brazil is described in paper 4.64. Whilst in paper 4.65 the authors look at the potential impact of distributed generation based on domestic scale fuel cells to the distribution companies of Brazil.
A probabilistic loadflow is compared with experimental results of increasing a CHP plant output from 2.7 to 5.2MW in paper 4.67. The authors from Poland observe that the loadflow gave more pessimistic voltage values than those measured but were useful in the decision as to whether to allow the greater output.