By Paul Eardley-Taylor and Nicholas Green (Standard Bank)
Increased electricity tariffs and historic, and potentially looming, power shortages have led to a widespread debate in the South African electricity sector. In line with the National Energy Regulator of South Africa (‘’NERSA’’) approved tariff increases, the increased electricity tariff that the country is already paying is projected to continue increasing for the foreseeable future. Some naturally argue “the tariff is too high” and that “it’s going to negatively affect the economy”. Clearly, increased tariffs mean greater monthly expenses for industry and for the residential consumer. However, a more in-depth analysis of the energy economics underlying the sector is argued necessary to determine whether the balance is appropriate.
Demand and Supply
South Africa’s Integrated Resource Plan (‘’IRP2010/Report’’) uses a top-down approach to forecast electricity demand for the country. The demand forecast is based upon the close relationship between electricity sales growth and GDP growth. The report assumes a moderate GDP growth trajectory of an average of 4.5% for the next 25 years, which one can assume to be reasonable. It additionally assumes a gradual shift from a traditionally energy intensive economy, both industrial and residential, to a less energy intensive economy, with energy efficiency steadily increasing over time until electricity demand is 35% below a ‘business as usual’ baseline by 2034. It is expected that a portion of these energy efficiency gains will come from specific Energy Efficiency and Demand Side Management (‘’EEDSM’’) initiatives. The likelihood, efficiency and effect of these initiatives would need to be interrogated further.
This assumption of a substantial reduction in energy intensity goes against one of the key pillars of South Africa’s industrial policies (maintained under successive governments); to promote industrialisation based on mineral beneficiation and manufacturing sector output (inclusive of re-industrialisation). Any gains in efficiency, and thus reduction, in energy intensity derived from the price elasticity and efficient technology effects, may therefore be partially eroded by any increases in industrial activity, especially the expansion of platinum, manganese mines and chrome smelters that are planned, and for which South Africa has comparative advantages in terms of reserves.
The specific break-down of the suggested 35% demand reduction due to EEDSM is not clearly explained in the IRP2010, it simply states that it is attributable to efficiency gains from the reduction in energy intensity. However, some conceptual drivers have been highlighted:
- Demand Side Management initiatives, largely driven by Government bodies. Many such initiatives, although only in their infancy, have greatly fallen behind projections and we can reasonably assume will continue to do so.
- The price elasticity of demand for electricity is expected to drive the majority of the EEDSM reductions. This is likely to lead to behavioural changes especially in residential users by a greater extent than that of industrial users.
- Efficient technologies. This broad category would also overlap partially with the above two drivers. For example, smelters in South Africa have historically been less efficient and are now installing waste heat-to-power and other cogeneration applications in order to improve efficiency
It is reasonable to assume that the price elasticity of electricity may infer that if the tariff continues to increase, electricity consumption may decrease. The debate that will have a greater impact on the demand for electricity, is around whether the Energy Intensive Users (‘’EIUs’’) will consume less, even with the tariff increasing – in essence are they sensitive to operating cost increases? The amount of EEDSM that are implementable in EIUs is likely less than the Department of Energy (DoE) is expecting per the IRP.
Albeit the price elasticity of electricity will have an impact on EIUs, one cannot overlook the importance of security of supply for these users and could even potentially outweigh the impact of price. As is widely acknowledged, South Africa does not have sufficient installed electricity capacity, negatively affecting the security of supply, and hence stunting potential growth and dissuading the much needed and important Foreign Direct Investment (‘’FDI’’).
As with any investment decision, FDI decisions are made around return on investment. In comparing South Africa to other developing countries/economies; such as India, China, Brazil, Eastern Europe and the Rest of Africa, it is argued South Africa not only needs to be price competitive in the resource sector – mining, ferrochrome smelting etc, in order to incentivise investment, but the country needs to be able to ensure the security of supply. Disruptions to supply, as South Africa experienced in 2008, can lead to the loss of billions of Rands in forgone production and GDP. With electricity being a large cost component in this sector, as well as the energy intensity of the associated industry, the electricity tariff and security of supply have a great impact in an investment decision of this nature.
The South African government therefore ideally needs to increase the available capacity, whilst keeping the pass-through of cost increases as low as possible. This can only be achieved by spreading the cost of new generation over the longest time possible. Electricity generation assets have a long lifespan, thus making this possible if there is the capital available to fund the new build.
Fundamentally, the decision to build South Africa’s two new, somewhat controversial, mega coal-fired power stations, Medupi and Kusile, should potentially have been the outcome of a comprehensive government policy, such as an IRP. However, the decision to build the power plants was made outside of a robust and conclusive publicly debated policy document process. Electricity tariffs have been increased largely to fund higher operating costs (coal, mid-life maintenance) and will further be increased in order to fund the Medupi and Kusile build programmes. However, the IRP2010 debate only started some two and a half years after Medupi’s start of construction, likely inferring the economic analysis at the time imperfect.
In order to better understand the themes touched upon already, a further breakdown of the tariff increase drivers is necessary. Eskom’s tariffs have historically been amongst the lowest in the world. Real tariffs were broadly flat to slightly declining for much of the mid 1980s-2000s. Since June 2007, and due to its New Build Programme (‘’NBP’’), increased maintenance capex and fuel / operating costs has led Eskom to request high nominal and real price increases, which NERSA has “settled” for at lower rates than requested – albeit representing large increases.
Blended wholesale tariffs have significantly increased in recent years, from R0.18/kWh in 2006/2007 to a current R0.55KWh (2011/2012).
Per NERSA’s most recent approval, tariffs will be R0.66 kWh by 2012/2013. Based upon Eskom’s presentation to Parliament (May 2010), it assumes for the next MYPD process, 25% annual increases in each of 2013-2014 and 2014-2015, and a 6% increase in 2015-2016. This would take average wholesale tariffs to R 1.09 kWh by 1 April 2015. On top of this, most end consumers also pay additional charges to help fund municipal distributors (e.g. CityPower).
Even with these highly publically contested NERSA-approved electricity tariff increases, Eskom stated it was exposed to a R190billion funding gap (although this number has not recently been updated). It is thus likely that Eskom would require further increases by 2017. The approved increases in tariffs will make alternative technologies, specifically wind power, increasingly price competitive.
There are, however, ongoing debates that the economy is becoming less energy intensive, due to price elasticity – the tariff is still increasing. The consequence is the EIUs do not want these tariffs to increase disproportionately to that of countries competing for the same FDI that would potentially flow into the South Africa borders; such as India, Kazakhstan, China, to name a few. It is likely that they understand that an increase in the tariff is justifiable and required to ensure security of supply, but it is the rate at which it is increasing that is being debated, balanced with the associated risks to security of supply.
Further to increasing electricity tariffs, National Treasury (‘’NT’’) published a discussion paper on proposed carbon taxes in December 2010. The government believes that a carbon tax appears to be the most appropriate mechanism to reduce Greenhouse Gas (‘’GHG’’) emissions, creating incentives for emissions reduction at the least cost to the economy.
While it would not guarantee a fixed quantitative reduction in such emissions over the short term, NT purport that a carbon tax set at an appropriate level and phased in over time would provide a strong price signal and certainty to both producers and consumers, acting as an incentive for more environmentally friendly behaviour over the long term. The Long-Term Mitigation Scenarios report (2007) and the National Climate Change Response Green Paper (2010) for South Africa recommends the use of market-based instruments, specifically carbon taxes, to induce behavioural changes that contribute to lower GHG emissions. NT believes the pricing of carbon through an appropriate tax will create the necessary incentives to change behaviour and achieve emission reductions at least overall cost to the economy.
They further argue that practicalities favour a proxy tax base based on the carbon content of fuel inputs. It would appear that they believe, a tax of R75 per ton of CO2, with an increase to about R200 per ton CO2 (at 2005 prices) would be both feasible and appropriate to achieve the desired behavioural changes and emissions-reduction targets. Ultimately, any introduction of a carbon tax will boost the competitiveness of renewable energy technologies in South Africa due to no CO2 charge being passed through. Indicatively, Standard Bank believes the Discussion Paper carbon taxes may increase South African electricity prices by R 0.12 – 0.14/kWh, although we note a revised paper will be issued shortly by NT.
Having looked at the IRP and the country’s tariff path increase implications, one also needs to understand the political landscape within which the implications have an effect. The DoE faces mounting pressure with regards to tariff increases and insufficient capacity. It is tasked with ensuring the electricity supply of the country, but ultimately does not set the price of electricity and thus has limited room to operate.
Moreover, Government has provided significant guarantees – R 350 bn guarantees plus R 60bn shareholder loans – R410bn in total – required to fund the NBP, due to Eskom’s internal inability to fund the NBP on its own Balance Sheet. This has provided additional pressure on the fiscus, when one takes into account that there are more power consumers than actual tax payers. As the Government’s R410bn fiscal contribution needs to be spread across the small tax base, and cannot be spent on the entire country, the additional strain on the fiscus, through existing commitment to social expenditure infers that the increased generation capacity expenditure pressure needs to be passed onto the user through tariff increases – sooner or later.
Government therefore has the onerous task of striking a critical balance between their own fiscal budget, the existing capacity of Eskom’s balance sheet and the rate of borrowing for the Medupi and Kusile builds.
Additionally, the expected increase in the price of coal is driving up the electricity tariff – the IRP2010 modelling takes into account a 20% increase in the cost of coal. So much so, that even without the increase in generation capacity, it is expected that the increase in the coal price, coupled with the increase in operations and maintenance (O&M) on the existing and return to service plants, will drive the tariff up to at least R0.78 kWh (real) in the medium-term. Assuming that the overall cost of Medupi and Kusile has now been finalized, the impact of the expense on the fiscus effectively comes down to the timing of the guarantees; the loans being raised by Eskom (against the guarantees) and the associated tariff increases.
Having taken into consideration the impact of new build power generation capacity on the fiscus and the tariff, we need to consider where new capacity would come from, if not developed and paid for by the government, and how much it would cost.
The Impact of Renewable Energy
Essentially, we should not compare the cost of new capacity to that of the existing capacity. Due to the historic legacy of Eskom, the existing capacity is old, and produces electricity at tariffs cheaper than almost all other countries. Moreover, Eskom has committed to a New Build Programme which will increase future tariffs.
This is where the IRP comes in and its allocation to renewable energy technologies enters the market. South Africa is presently one of the world’s most exciting renewable energy markets, adopting the technologies ‘late’ in comparison to more developed markets, but with a high expected growth rate over the next 10 to 20 years; a mechanism that will aide in addressing the electricity supply shortfall that South Africa faces.
The IRP2010 is, in effect, the South African Government’s 20 year energy sector master-plan, issued for public consultation in October 2010; Cabinet approved on 16th March 2011 and promulgated on 6th May 2011. It states that going forward, 42% (17.8 GW) of South Africa’s new generation capacity is proposed to come from renewable energy, per the below:
§ Solar PV: 8,400MW
§ Concentrated Solar Power (CSP): 1,000MW
§ Wind: 8,400MW
In order to make a reasonable cost comparison, we would argue that Solar PV and CSP technologies should be benchmarked against that of a diesel-fired peaking plant (SA’s current reserve margin). The extremely high variable running cost associated with peaking plants can be mitigated against with the installation of Solar PV and/or CSP plants, which would yield a similar/equivalent load generation.
Standard Bank believes that the tariff charged by new wind farms, per kWh, could achieve parity with Eskom’s blended wholesale tariff as early as 2015/2016 and that new build Solar PV could achieve the same by 2018/2019.
A Renewable Energy Feed-In Tariff (‘’REFIT’’) was the initially planned route to market for renewable energy. This has since been replaced by the Independent Power Producer Procurements Programme (‘’IPPPP’’), which was released on 3 August 2011. The IPPPP is to account for the initial 3,725 MW of renewable energy in South Africa, of which 91% relates to Wind and Solar. The programme specifically relates to renewable energy Independent Power Projects (IPPs) and incorporates a competitive price bidding process, for which the revised tariff is set per technology, per the below diagram.
The IPPPP includes the following technologies:
§ Onshore Wind,
§ Solar Photo-Voltaic,
§ Concentrated Solar Power,
§ Small Hydro,
§ Landfill Gas (LFG), and
The IRP2010 requires 300 MW Solar PV per year from 2012 to 2024, 400 MW of Wind from 2014-2023, 200 MW CSP by 2015, followed by 100 MW per year through to 2025. The IPPPP aims to kick-start the process of reaching the IRP2010 targets, with the allocations represented in the diagram below.
Eskom will be the buyer of the power produced through a standardised PPA, acting through its Single Buyer Office (SBO). The government will provide support for PPA payment obligations through the Implementation Agreement with DoE.
Standard Bank has been pleased to be involved in multiple debt and advisory mandates for the IPPPP’s Bid Date 1 (“BD1”) and expects to be further involved in BD2. We are beyond doubt that, assuming closure; IPPs will work in South Africa. Their ability to assist the government in striking the balance required is paramount to the success of increasing the country’s installed capacity, and keeping the tariff down in the long-term.
It is the intention of the DoE to update the IRP on a bi-annual basis, therefore allowing for changes in technology allocation. This has a meaningful potential outcome, whereby the current allocation may become differently weighted going forward. E.g. more Solar PV, less wind, or increasing the amount of CCGT and decreasing the allocation towards nuclear per se.
The nuclear allocation in the IRP2010, as well as the small allocation to CCGT/OCGT may well become a prioritized discussion point for the DoE going forward. The existing 9.6GW nuclear allocation was finalized and included in the IRP2010 prior to the subsequent accident at the Fukushima Daiichi plant in Japan; which has since raised numerous questions and spurred on debates globally as to the safety of this powerful energy source. Most observers’ question whether a new nuclear power station will be built in South Africa, albeit the decision will ultimately sit with the DoE whether it is to be included in future iterations of the IRP.
Taking note of this, we now turn our attention to the potential for Gas-to-Power in South Africa and the surrounding countries. Per the below schematic, due to the increasing gas discoveries and advances in gas extraction technologies, we foresee greater focus on Gas in the South African market and in future iterations of the IRP.
There are currently multiple real-time choices surrounding the supply of gas to South Africa:
1. West Coast – Substantial gas deposits have been found off the west-coast of both South Africa and Namibia
2. East Coast – Mozambique and Tanzanian Gas
3. Shale Gas – The US EIA estimates the potential for 485 TCF of Shale Gas in the Karoo basin
The above fuel-source options can be used in multiple configurations. For example, they could provide a secure supply of gas to the existing OCGT plants at Ankerlig and Gourikwa, as well as fuel new CCGT plants or, for that matter, the Mossel Bay GTL plant. Additionally, the carbon emissions associated with the burning of natural gas is lower than that of our existing energy infrastructure, lowering the impact of carbon taxes and thus the tariff.
The potential clearly exists, but the window of opportunity to secure this supply is not wide (for example, 2-3 years to agree commercial terms). The Mozambican and Tanzanian gas supply could solely be sold to alternative off takers in the global markets due to oil-based contracting structures in place for LNG. SA power players as is known sell at Rand denominated prices. Should South Africa decide to take advantage of the opportunity available to it, it should place an increased impetus on the development of a more robust gas infrastructure It would naturally be critical that we would need to get it right first time round to ensure that the correct balance can be struck.
Accordingly, South Africa faces a difficult balancing act in relation to its power sector (as policy-makers indeed disclose). Official projections (IRP 2010) note current and projected shortfalls and the need to diversify generation sources and increase tariffs. However, implementing such increases raises challenges in a open society facing multiple social and economic trade-offs. At the time of writing, South Africa appears to have made a good start in addressing one policy objective – renewable energy, but numerous other challenges must be successfully faced to ensure policy success in order to facilitate continued economic growth and achievement of industrial policy objectives.