South Africa will experience routine electricity blackouts in a few years unless new electricity policy and investment decisions are formulated and implemented this year.

This is the inexorable conclusion that emerges from scenario and modelling exercises undertaken separately by the National Electricity Regulator, Eskom and large energy-intensive industries.

Growing electricity demand will outstrip existing national supply capacity next year or the year thereafter, assuming a prudent reserve margin to allow for maintenance and unscheduled plant shutdowns.

Hydro-electricity imports, mainly Cahora Bassa in Mozambique, will provide respite for about another year. Thereafter, we need further generation capacity or significant energy savings and demand-side measures.

Eskom has started re-commissioning old moth-balled coal-fired power stations to meet this challenge. Camden, the first plant, will be relatively easy to re-commission and work has commenced. Grootvlei will be more difficult and Komati, the last plant that Eskom plans to re-commission, will be the most uncertain and expensive.

If successful, these old generating stations will give us a breather until around 2008. And then we need new generation capacity.

2008 might seem years away, but investment decisions, environmental impact assessments, plant construction and commissioning take many years. For a hydro-electric or pumped storage scheme, this could take ten years. A coal-fired power station could take six years or more, and gas turbines - two to four years.

If our economy grows faster, or we are not able to implement effective demand-side measures, new power generation capacity might be needed even earlier.

Government is aware of this situation. The President confirmed, in his state of the nation address in parliament in May, that a tender for new capacity will be awarded early in 2005.

The Department of Minerals and Energy has appointed technical advisors to prepare and manage this tender. However, their work schedule indicates that the contract with a new Independent Power Producer will only be concluded early in 2006, and this will only happen if the bid manages to comply with National Treasury's Public Private Partnership regulations. The DME will have to show that Eskom cannot build a new plant more cheaply - an interesting possibility given Eskom's competitive cost of capital and the potential for transfer-pricing with its current portfolio of extremely low-cost generating plant.

Given these tight time constraints, it is not unlikely that we shall have to resort to buying, on an emergency basis, a series of highly expensive, paraffin-burning open-cycle gas turbines.

There is a dangerous assumption that the current tender process for new generation capacity answers concerns about supply security. It does not.

The challenge is not only to manage the current tender process within tight time-constraints. We need to make decisions this year about procuring much more capacity than the approximately 1000 MW anticipated in the current tender.

A likely planning scenario indicates that this year, 2004, we need to make investment decisions on a new pumped-storage scheme, a new pulverised coal-fired plant and a green-field coal fluidized-bed combustor or a combined-cycle gas turbine. In short, we need to start placing orders for a range of new power plant. In ensuing years we shall need to continue to order new plant.

These challenges raise the question of whether a part-time committee of government officials, assisted by consultants, is the most appropriate and sustainable mechanism to continue to procure new power? It also provokes debate about what market structure is appropriate to encourage the most efficient and cost-effective investment decisions?

Following the 1998 While Paper on Energy Policy, and a number of subsequent studies, Cabinet decided, in May 2001, to restructure the power sector by unbundling Eskom's electricity transmission division into an independent company and selling-off 30% of Eskom's generation plants. New capacity would be provided by private investors and an electricity trading market would be established comprising a power exchange and a parallel market for bilateral power contracts and financial hedges. None of this happened.

What is emerging is a quite different market model. In her budget speech, the Minister of Minerals and Energy stated that "the state has to put security of supply above all and above competition especially". The Minister of Public Enterprises has indicated that Eskom will not be privatised and that a strong state-owned utility is important for social and economic development.

Eskom is thus likely to continue to dominate the market. It may even be permitted to build new generation plant. Private sector investment will be permitted only on the margins in the form of Independent Power Producers. They will sign long-term power purchase agreements with Eskom (or with an independent transmission company or system operator, if these are eventually separated form Eskom).

Government will now need to clarify whether the emerging market model for the electricity sector is its preferred model or is merely a temporary measure to secure emergency supplied. This is not a trivial question - for it strikes at the heart of the cost and efficiency issues in the power sector, and will have long-term consequences for electricity prices in this country.

Few remember the controversial electricity price-hikes by Eskom in the late 1970s and 1980s when it made investment mistakes that resulted in huge unused power generation capacity. History demonstrates the potential weaknesses of the old industry model where state-owned monopoly utilities simply pass the costs of poor investment decisions to consumers.

The current tender process is also full of risk. A small number of officials and technical advisors will decide how much new power is needed, using which fuel sources, when and where. While a degree of (once-off) competition might be possible through the tender bids, long-term power purchase agreements could tie-up non-competitive electricity prices for decades.

Plans for a new market structure, where investors have to compete to sell their power in a power exchange or a contract market, have been sacrificed in the face of security of supply concerns.

Periods of supply uncertainty and shortages are never a good time to design and implement new competitive market structures. The long period of large capacity surpluses that provided a window of opportunity for major reform has disappeared. Now we have to patch the current system and prepare for the future.

The default IPP/ single-buyer model that is emerging now requires the establishment of a robust and sustainable institutional structure (probably best attached to the power system operator) that will be responsible for long term planning, security of supply and procurement of generation capacity.

We can avoid future black-outs. But we need to act now.