In the last few years, the Power Generation Sector in Singapore has been struggling along with too much production capacity and too low prices. The Supply Cushion1 has been over-expanded leaving a significant part of the production arsenal unused.
We have previously written about the Loss-making Power Generation Sector in Singapore. Singapore’s Market Model for electricity determines the price of electricity at the Short Run Marginal Costs (SRMC) of a CCGT power plant. The problem with these markets is that one is not covering its fixed costs on marginal plants at the SRMC. And, CCGT power generation plants produce most of the electricity in Singapore.
Some of the fixed costs are covered by the Vesting Contracts, in particular through the LNG Vesting Scheme. In an effort to diversity Singapore’s energy supplies with LNG, the LNG Vesting Scheme was introduced to incentivise new investors to use LNG as fuel for new CCGT power generation. It compensates any revenue shortfall up to the Long Term Marginal Costs (LTMC) which includes fixed costs. This biased investment signal went beyond the price signals of Singapore’s Market Model and resulted in new CCGT capacity being build. If it was left to the pricing signals, little or none of this capacity would have been installed.
Unfortunately, these newly build power plants resulted in an overcapacity of CCGT capacity. And consequently, this reduced the Uniform Singapore Energy Price (USEP) to the SRMC of CCGT plants and eliminating profitability for the Generation Companies (Genco’s).
The question is whether it is likely that price signals from Singapore’s Market Model will correct the CCGT overcapacity through divestments or mothballing of generation capacity. Is it possible for the sector to organically loose overcapacity to restore profitability to the sector?
How much Capacity is Overcapacity?
The Supply Cushion indicates the level of available but unused capacity. Some of it is required to guarantee supply adequacy which can be called up when required. Historic data of the Supply Cushion shows three distinct phases:
Figure: Supply Cushion 2003 to June 2017
Phase 1 covers the period until early 2009. With the supply cushion just over 37%, USEP prices were on average a little below Vesting Prices derived from the Vesting Contracts. And, as a result, during this period on average the Vesting Contracts made the Singapore Consumers pay to compensate the Genco’s up to the LTMC of CCGT on 55% to 65% of the electricity demand.
Phase 2 starts in 2009 and last until 2013. An increasing electricity demand, following the completion of a number of infrastructure projects, and a simultaneous reduction in available capacity reduced the supply cushion to 31%. This reduced supply cushion together with tightness in the supply of natural gas2, forced peak-demand into more expensive steam turbine capacity. And the higher SRMC of this capacity resulted in higher USEP which, on average, exceeded Vesting Prices.
Figure: Available Generation Capacity 2003 to June 2017
In Phase 3 the supply cushion surges to over 42%. The commissioning of the LNG terminal came along with more than 3,000 MW CCGT capacity fuelled by new LNG supplies. In the same period, electricity demand growth remained subdued. The steam turbine capacity is now out of reach for peak-demand and the Singapore’s Market Model determines the USEP consistently on the CCGT SRMC. Although the Vesting Contracts still compensate the Genco’s between 40% and 25% of the electricity demand, most of this compensation is dedicated to the new LNG fuelled CCGT capacity.
An energy-only market model, like the Singapore’s Market Model, only functions an economically sustainable equilibrium under certain conditions. It is critical that the supply consist of different types of generation facilities, each with its own distinctive marginal costs. Normal price signals would prevent an investor from adding new CCGT capacity which pushes peak-demand out of the higher cost generation like the steam turbine capacity. Because this would reduce the USEP and therefore the profitability of the new CCGT capacity (and all existing CCGT capacity at the same time).
We estimate that a reduction of the supply cushion to 36% would restore this equilibrium. This implies that approximately 1,800 to 2,000 MW needs to be retired or repositioned from Singapore’s total generation capacity.
How to retire 2,000 MW?
One might think that the obvious candidates for retirement are generation facilities which are currently not required for peak-demand. This would place the highest SRMC plants, the steam turbines, on the top of the list. However, this would not help the purpose of the retirement. The terms of pricing by Singapore’s Market Model remain the same and the USEP is still determined by the SRMC of CCGT alone.
It is therefore essential that CCGT plants are retired. This would push peak-demand back into steam turbine capacity and USEP into the steam turbine’s SRMC. As the LNG Vesting Scheme subsidises the new LNG fuelled CCGT up to its LRMC, there will be no incentive to retire any of plants covered by the LNG Vesting Scheme.
The CCGT plants fuelled by pipeline gas and owned by the ‘original’ incumbent Genco’s are the most likely candidates. However, the Genco’s are caught up in a complicated situation. If one of the Genco’s sacrifices part of its arsenal, it provides the benefits to others. Also, the required retirement is soo large, it would threaten the existence of a single Genco when executed single handed. At the same time, when nothing is done, they all remain loss-making. There seems no organic way to resolve this impasse.
Retirement of up to 2,000 MW can only be achieved with a coordinated effort by the incumbent Genco’s. When each retire a part of their CCGT capacity they could all benefit from it. The Energy Markets Authority (EMA) could help this by redistributing of the benefits of the USEP increase by other producers as compensation for the loss of share of the retired plants. Whether it is politically acceptable that modern CCGT is retired to increase prices for electricity is to be seen. It is all-in-all a complicated affair.
Can 2,000 MW be repositioned instead?
Instead of retiring available capacity, one could expand the demand for electricity. CCGT capacity is repositioned for this demand and the supply cushion is reduced. The obvious method is to expand the Singapore market beyond its borders though export cables to Malaysia or Indonesia. Currently a submarine cable between the Senoko Power Station in Singapore and the Sultan Iskandar Power Station in Johor in Malaysia is used as back-up capacity in case of power outages.
Excess capacity in Singapore can be a new supply competing with planned new capacity abroad. An Export Agency could build subsea cables and buy 2,000 MW of power for export at SRMC prices. This would allow the Singaporean market to restore to an economically sustainable equilibrium between supply and demand. The capacity available, other than for export, will be priced at a USEP which provide long-term returns for the power generation sector.
The sale of the exported electricity could generate surplus revenue of its cost price of SRMC. This can be used to finance the investments in subsea cables or provide additional income for Singapore.
However, the idea that prices for electricity in Singapore increase as a result of exports to neighbouring countries might also face difficult political hurdles.
Do Pigs Fly?
Is it likely to happen that up to 2,000 MW is retired or repositioned for purpose to increase prices in favour of the Genco’s but paid for by Consumers? In the short-term probably not. The EMA has been extremely careful not to take any responsibility for the current situation. The overcapacity is too large for the incumbent Genco’s to resolve organically.
In the long-term, the economic viability of the Genco’s is at risk. One way or another, something will happen.
1. Supply cushion is the ratio between (a) the supply and demand gap (i.e., the difference between available capacity and demand) and (b) available capacity. This index measures supply adequacy. It indicates the level of unused capacity that was offered but not scheduled, and could be called up if required. The available capacity refers to the total amount of capacity made available by all generation registered facilities minus reported outages. Demand refers to the demand forecast by the PSO used to determine the real-time dispatch schedule for energy.
2. The EMA imposed a policy prohibiting the import of new natural gas supplies through existing Malaysian and Indonesian pipelines. This policy was meant to support the demand for LNG and underpin the development of the LNG terminal. Prior to the commissioning of the LNG terminal, this policy caused tightness in the supply of natural gas.