Security of electricity supply in South Africa

December 19th, 2014, Published in Articles: EE Publishers, Articles: Energize


A confidential report to the Department of Public Enterprises, dated July 2006, on security of electricity supply in South Africa, written by Duncan Wilson and Ivan Adams, has come to the attention of EE Publishers.

Although marked confidential, the report has some historical significance, having been submitted some 18 months prior to the loadsheedding and blackouts of December 2007 and January 2008, and is considered to be in the public and industy interest. The executive summary of the report is therefore published here, with a link to the full report for those interested in further details.

Review of security of supply in South Africa

A report to the Department of Public Enterprises

by Duncan Wilson and Ivan Adams

July 2006

Click here to download the full report

Executive summary

In South Africa, security of electricity supply is one of the most important issues facing the  electricity industry  currently,  its  customers  and  Government.  Key  aspects  of electricity supply security are the availability of adequate generation capacity to meet customer demand at any time and a secure and reliable transmission system to deliver power to all regions of the country. Maintaining a secure electricity supply is essential for any developed economy. However, determining the appropriate level of supply security is a trade-off between the costs involved in improving power system reliability and the losses to the economy and customer welfare associated with power outages. This study makes recommendations and provides guidance as to how that trade-off can best be achieved.

Following the difficulties at Koeberg nuclear power station, the electricity supply and demand balance is extremely tight, particularly over the current winter peak demand season. Even with a resolution of the Koeberg problems, the national supply/demand balance is likely to remain tenuous during the next few years. Given Eskom’s key role in planning and managing supply security, it is therefore appropriate for the Department of Public  Enterprise  (DPE)  to  develop  a  position  on  security  of  supply,  for  both  the immediate short-term following the Koeberg incidents and the medium to longer-term. In developing this DPE position, it is proposed to engage with Eskom, DME, NERSA and other relevant parties that have a responsibility for electricity supply security and energy policy.

Internationally, the importance attached to power system security has increased significantly in recent years, following large-scale power outages in North America and Europe. In many countries, the traditional approach of leaving security to be managed by a utility has been replaced by a transparent and inclusive process, with Governments and Regulators taking a pro-active position. In those markets where competition does not exist or is not extensive, an explicit generation security standard is usually set, as a basis for future planning. In those markets where there is extensive competition in generation and retail, Governments and Regulators are clearly concerned about the ability of the market to deliver adequate security of supply and are putting in place measures to ensure adequate security. International experience also shows that maintenance and asset management are crucial to avoid increases in the   number   of   concurrent   fault   events   that   would   risk   more   and   larger interruptions to customers’ supplies.

The security of an electricity system is affected by all elements of the supply chain – generation, transmission and distribution. The main focus of this study is on generation and transmission.

The  adequacy  of  generation  capability  depends  upon  such  factors  as  the  installed capacity, unit size, plant reliability, demand forecasting error and the shape of the load curve. The Reserve Margin is a deterministic criterion, which provides perhaps the simplest  available  measure  of  system  security.  However,  it  does  not  take  explicit account of the fact that security is dependant upon underlying factors such as the size of individual generating units and the relative reliability of generating units on a system.

Eskom has recently stated that a reserve margin of 15% to 25% is the desirable range required to meet Eskom’s obligation to supply (OTS), although this OTS is not defined explicitly.  However,  this  level  of  plant  margin  is  somewhat  higher than  has  been considered appropriate  by Eskom in the past and it is unclear what has driven this change in policy.

As  the  reserve  margin  on  a  system  is  increased,  the  probability  of  failing  to  meet demand as a result of inadequate generation will fall. Due to the probabilistic nature of security of supply, it is only possible to provide an expectation of failing to meet demand. As  load  grows,  the  reserve  margin  is  eroded,  until  such  time  as  new  plant  is commissioned. The decision as to when new plant should be commissioned depends upon what reliability of supply is deemed to be appropriate. Increased reliability implies greater investment costs while reduced reliability results in an increased expectation of power shortages, which have an implied cost to customers. Appropriate reliability criteria are derived by balancing these two factors.

The construction time of new power plant has a dramatic impact upon the level of uncertainty  involved  in  generation  planning.  The  longer  the  construction  period,  the greater the uncertainty, due to such factors as demand forecasting error, performance of existing  generating  units  and  uncertainty  over  the  commissioning  dates  of  new generation units. For example, the decision as to whether to build a combined cycle gas turbine (CCGT) plant, with a relatively short construction time, or a coal fired plant, will have a significant impact upon the required plant margin in South Africa.

A key criteria used by Eskom for determining the need for additional generation capacity is the “Cost of Unserved Energy” (CoUE), which is assumed to represent the value to customers of system security. This approach, although widely used in the past in certain countries such as the USA, is now recognised as not being an adequate basis for determining  power  system  security,  as  has  been  illustrated  by  the  response  of customers to the recent power outages in the Western Cape.

The overall responsibility for supply security in South Africa lies with the Department of Minerals and Energy (DME) but in the short-term (to 2008) Eskom has an obligation to meet the need for additional capacity. However, there is no agreed basis or standard for the level of supply security to be provided. It is recommended that an unequivocal security standard should be established.

Eskom is currently responding to a critical shortfall in generation capacity, partly due to the Koeberg problems but also to the fact that over the past few years the reserve margin has fallen to record low levels compared with the historic situation.

There is now an urgent need, first to determine a suitable security standard, in order to define what level of reserve margin should be maintained and second, to clarify the responsibility for meeting that security standard, given the Government policy that new generation capacity shall be provided 70% by Eskom and 30% by the private sector.

The supply problems during 2006 have highlighted, not only concerns about generation capacity  in South  Africa  but also concerns  about  the adequacy  of the transmission system to deliver power to all of South Africa’s regions. It has become apparent that, although the transmission system is generally designed to be resilient to a single circuit outage, the transmission system does not meet this criterion in all regions. This is as a result of the economic criteria used by Eskom in order to justify transmission system augmentation, which are based, in part, upon the Cost of Unserved Energy (CoUE). The result of this approach is that in some regions of South Africa where demand exceeds the local generation capacity, the security of supply is lower, due to the limitations of the transmission system, than in those regions where the generation exceeds the demand. Effectively, customers in some parts of South Africa have a sub-standard supply security, as a consequence of a perceived value of supply interruptions, i.e. the CoUE, which has not been subject to rigorous validation.

A key element of system planning, both in the short-term and longer-term, is developing a reasonably accurate forecast of the demand. This study has revealed a number of areas where the current approach to demand forecasting could be improved, at relatively low cost. However, the main finding of the study, with regard to demand forecasting, is the need for a truly integrated approach. At present, a theoretically integrated approach exists within the National Integrated Resource Plan (NIRP), under the jurisdiction of NERSA. However, in practice it is apparent that the NIRP is largely irrelevant to Eskom’s planning, which is based on its own Integrated Strategic Electricity Plan (ISEP). It is appropriate that planning should be under the control of an independent body and there are  aspects  of  the  new  approach,  adopted  within  NIRP3,  that  are  advantageous. However, the current NIRP approach fails to take sufficient account of the considerable expertise on demand forecasting that has been developed within Eskom. In addition, municipalities, as retailers to around one-half of the customer load in South Africa, need to be more fully involved. There is strong case for the two separate planning processes, ISEP and NIRP, to be brought together in a fully transparent process, open to all parties and managed on a consistent and regular basis.

A major input to the demand forecast is the assumed level of economic growth in South Africa.  It  has  been  argued  by  Eskom  that  Government  targets  for  growth  may  be optimistic but recent figures suggest that GDP growth is currently strong and there are grounds for remaining optimistic about the future. Currently, however, there is lack of agreement, within the electricity industry, about the assumed level of economic growth  that  the  industry  is  preparing  to  meet.  This  is  not  a  good  basis  for planning.

The problems experienced with the availability of the Koeberg power plant have demonstrated very clearly that, with the generation and transmission infrastructure that is currently available in South Africa, the loss of a single 900MW generating unit at the Koeberg station for any extended period during the winter months results in Eskom being unable to meet customer demand in the Western Cape. This illustrates clearly that the security of electricity supply in South Africa is currently at risk from a single (albeit low probability) event.

Over the past 10 years the reserve margin has fallen very significantly as a result of growth in electricity demand of around 3% per annum (which equates to approximately 1,000MW of additional peak demand each year) and the very limited amount of new generating plant that has been commissioned. However, the monitoring of reserve margin is inadequate and inconsistent. There is no agreement between NERSA and Eskom on whether to include demand-side management (DSM) measures as part of demand or as a supply-side option and there is lack of clarity concerning the distinction between the reserve margin adopted for long-term planning purposes and that used for short-term operational purposes. It is recommended that these anomalies should be removed and that a consistent, transparent and regular monitoring system be established to track the reserve margin and the availability of generating plant and the transmission system.

Eskom’s plans for generation capacity expansion (ISEP) concentrate on the long- term  supply  position  and  do  not  focus  on  alleviating  the  shortfall  in  supply security   in  the  short  term.  If  the  return  to  service  of  mothballed  plant,  the commissioning of any of the new gas turbines run behind schedule or the forecast level of Demand-Side Management (DSM) fails to materialise, the capacity situation in the short term will become tighter. Contingency measures for dealing with such possibilities should be in place.

Steps  could  be  taken  now  to  improve  the  short-term  position.  For  example, additional  open-cycle  gas  turbine  (OCGT)  plant  could  be  procured,  possibly  as extensions to the existing tenders by Eskom and DME. The advantages of OCGT plant are flexibility (including the possibility of later conversion to run in a combined cycle mode), ability to locate in regions where generation is in deficit (thus adding to security in those regions) and the relatively short timescale for construction. The high cost of fuel is a factor to be considered but not to cause rejection without taking account of other factors.

Although imports and exports are relatively small, compared with the total electricity demand in South Africa, the levels of imports and exports are significant in comparison with the reserve margin. Thus, there is a need for a stronger focus on imports and exports, particularly with regard to the contractual commitments associated with both. It is also recommended that the risks to supply security, associated with imports, be monitored on a probabilistic basis.

In all of the other markets reviewed as part of this study, there is significant information  in  the  public  domain  concerning  security  of  supply  and  reserve margin. By contrast, in South Africa, almost all of Eskom’s documents on security of supply  and  reserve  margin  are  either  unpublished  or  specifically  deemed  to  be confidential.  Whilst  Eskom  may  argue  that  their  ISEP  results  are  confidential,  the argument is invalid, particularly for a dominant state-owned entity not in competition with other utilities. International experience shows that in both competitive and monopolistic markets, the type of information produced in Eskom’s ISEP is deemed to be public domain information, in order to ensure that all interested parties are aware of the plans, issues and options under consideration. It is also likely that a more open process may have revealed the serious disjoint that exists, between the long-term planning process and the process of delivering the required capacity additions. The lessons are that, where key decisions are being made, it is vital that the plans and the associated assumptions have the widest possible exposure to enable alternative ideas and options to be considered. Eskom does not have a monopoly on good ideas.

It is therefore strongly recommended that Eskom’s ISEP results, together with the associated  procedures  and  policies,  should  be  published  in  full.  However,  it  is recognised that there may be circumstances where certain information may need to be kept anonymous for reasons of commercial confidentiality.

In addition to making the ISEP process fully available to the public, it is recommended that a more consistent approach be taken towards the updating of the ISEP demand forecasts and expansion plan options. It is recommended that a demand forecast should be updated at least annually; possibly at some point soon after the actual winter peak demand has been established.

It is recommended that the DPE should initiate discussions with NERSA, Eskom and all interested parties with a view to integrating the ISEP and NIRP processes into one seamless national process. It is recommended that the single integrated planning  process  be  managed  and  co-ordinated  by  NERSA,  with  Eskom  being responsible for agreed components of the plan. The outputs from this plan should be made public at least on an annual basis and more frequently if specific circumstances require it. In particular, in a period where the supply situation is tight, it is vital that a current plan is available for immediate implementation and that such information is in the public domain on a timely basis.

Click here to download the full report

Related Articles

  • Minor improvements to silicon crystalline PV cells increase efficiency
  • New series thermal imagers
  • Locally made inverters for new solar farms
  • Assessing legal risks in moving to the cloud
  • Seminars peak interest in pump engines