Vexed Matters in Ghana’s Power Sector: Take or pay versus take and pay – which way forward?
Ghana’s current excess generation capacity issue surfaced towards the end of 2015. But the discussion intensified following a study by a World Bank fact-finding mission in April 2016 dubbed Ghana IPP Guarantee Project. The Bank stated that every 100MW of power added to the grid costs US$40M in annual capacity payments. They projected excess capacity payments would reach US$450M by 2018, equal to 1% of gross domestic product (GDP), an expense the Government of Ghana (GoG) had no fiscal space for. The debt in the energy sector then was US$2.5B, with ECG as the main defaulter, causing a systemic disruption in the energy sector and in the general economy. The Energy Sector Levy Act (ESLA) was launched in 2015 to address the debt overhang but the problem persists.
However, there were some issues with the World Bank study. First, they included in their project-pipeline a couple of Independent Power Producers (IPPs) with doubtful paths to commercial operations. Second, they were overly optimistic about the commercial operation dates of certain IPPs. Finally, they reckoned overgeneration as temporary; Ghana will revert to generation shortfalls beyond 2022. That is now unlikely unless the Bank later revised its view of which this writer is not aware. However, their core conclusions were about right. In 2016 Ghana was coming out of a power supply crisis and some deemed it necessary for the activities of IPPs to be facilitated to avert a crisis relapse. But the Bank posited that natural gas for fueling existing plants was the issue and not generation capacity per se, and if gas were readily available, Ghana could have more than enough reserve margin.
The Finance Minister in his midyear budget review in July 2019 indicated Ghana has 5,083MW of installed capacity of which 2,700MW is utilised at peak load (he reaffirmed this in the 2020 budget in November 2019). Of this, 2,300MW has been contracted on take or pay basis costing GHS2.5B annually in excess capacity payments. And for gas, Ghana has contracted for 750 MMScf per day of which current need is only 250 MMScf and the excess capacity payments are estimated at between US$550M and US$850M per year. Excess capacity payments are a bane to GoG and in the 2020 budget a freeze has been placed on relevant state institutions in the energy sector regarding signing new contracts.
GoG declared its intension in the 2019 midyear budget review to switch from take or pay to take and pay contracts. Take or pay means whether the power producer is dispatched (asked to put power unto the grid) or not, it will receive payment for its available capacity to enable it to cover its fixed costs: debt service, fixed operating costs, and an agreed equity return. Some experts see this as a subsidy which increases the end user price of power and must not be allowed. But capacity revenue certainty incentivises investment in new and existing generation assets. Regarding this, Peter Terium, a former CEO of RWE, a large German power utility once quipped “you don’t pay the firemen only when there is a fire”.
Take and pay on the other hand commoditizes power and renders plant capacity irrelevant as the offtaker (buyer) only pays for the energy it buys based on the marginal cost of production. Volume of energy sales then becomes critical because without capacity payment and enough dispatch the producer loses. Revenue certainty is key to the bankability options and lowering the cost of debt of power producers but take and pay generally challenges the project finance structure. A bankable PPA is a long-term offtake contract with a creditworthy offtaker with sufficient tenor to enable repayment of project finance debt by providing an adequate and predictable revenue stream but the nature of take and pay generally limits this bankability option. The situation becomes even more challenging if it was not part of initial discussions but is roped into the process later. Consider this:
In July 2016 the Federal Government of Nigeria (FGN) announced a solar PV power programme with 14 IPPs to generate 1,125MW of solar power. Midway through the process, FGN requested the IPPs to cap their respective project costs at US$120M and take a feed-in-tariff of US$0.075/kwh, forcing the solar power market into a kind of ‘perfect competition’. There was push back. FGN introduced these conditions for undertaking to provide bankable Buyer Payment Security (BPS) to the IPPs in the form of put call option agreements, facilitation of partial revenue guarantees, and letters of credit. However, these conditions came late. Next, FGN sought to remove take or pay from the PPAs. In addition, it is seeking to discard the role of the IPPs and their financing parties as a condition precedent, requiring only FGN (as Buyer and Guarantor) to approve the BPS. This is recipe for misaligning interests as the BPS will need to be mutually bankable.
The programme is now in dire straits and certain options are being explored to save it from collapse. Some IPPs that had put up with the project cost cap and feed-in-tariff fiat are now primed to cut their losses rather than continue with the programme as price takers with potentially no take or pay and dispatch guarantees and with no revenue certainty regarding the volumes they will generate.
Be that as it may, take or pay and take and pay are two sides of the same dispatch risk mitigation coin; it is all about when and how they are applied. Dispatch speaks to energy that can be used on demand at the request of the power grid operator in line with market needs. Producers subject to dispatchable generation can switch their plants on and off and they can also change their production plans to meet a specific order and they are penalized if they cannot meet certain key performance indicators under a take or pay arrangement. Thermal power producers (oil, gas, and coal) fall in this category. This contrasts with renewable energy sources like wind and solar power not subject to dispatchable generation because they cannot be controlled as in dispatchable generation. But there are other renewables that are also subject to dispatchable generation without separate energy storage such as geothermal, biomass, and hydro.
The US Development Finance Corporation (formerly OPIC) and its affiliates have highlighted 10 risks to be mitigated by bankable PPAs. Dispatch risk tops the list. Both take or pay and take and pay are acceptable to lenders in mitigating dispatch risk but the latter works in a narrow sense. In 2014 the UK got approval from the EU Court to launch a capacity mechanism to reward power plants for their capacity availability rather than the energy they actually generate and other European nations were following the UK’s lead. But the EU Court years later rescinded its decision, throwing UK’s capacity program into chaos. Critics like Benedict de Meulemeester, managing director of international energy procurement consultancy E&C, aver that producers may lobby for capacity mechanisms but governments should not yield. Capacity payments drive up end user prices and they are an expensive solution for a non-existent problem, he argues.
Nevertheless, two things standout. First, take and pay works well with certain generation sources in a transparent auction market with a credible merit order dispatch. Because of the low-cost generation of renewables like wind and solar (with nominal marginal costs), they are among the first to be dispatched in any auction which “clears” the market at peak load. So, take and pay contracts for them is appropriate. Their marginal costs are covered by the market clearing tariff and because this market clearing tariff exceeds the individual tariffs of these renewables, the payments they receive cover their fixed costs too and leave them with profits. Second, because there are various generation sources, planning is required to properly balance the grid for take and pay to work well and so it may not be used to address a capacity problem which has festered over time in a market with nominal renewable sources and without total transparency.
Given Ghana’s limited power generation sources (despite the current excess capacity), and its underdeveloped auction market, it stands to reason that take or pay is the way forward for now. But to limit GoG’s payment exposure, there should be a transparent process with producers to ensure that capacity payment recipients are not “subsidized” to make supernormal profits. This is feasible if GoG collaborates with them to review their operating assumptions to optimise their cost structures. For example, to the extent possible, GoG may facilitate the refinancing of relatively expensive debts of producers such that it does not take any exposure on its own balance sheet. Further, account could be taken of the extent to which certain GoG initiatives may have benefited the power producers such as making gas available ahead of schedule thereby positively impacting the fuel costs of the producers. And if fuel tolling arrangements can be worked out as appropriate and incorporated in the respective contracts to mitigate the cashflow burden of the producers in funding the fuel cost themselves that will be great. GoG can then make a case to share the upside of all this with the producers in a tariff renegotiation in order to make savings on capacity payments. In addition to this, if GoG could look beyond ESLA and endeavour to go for a big bang approach to retire a big chunk of ECG debts to give producers cashflows, the notion of a win-win engagement will be more meaningful.
In the medium to long term, the promotion of small scale off grid generation solutions on a national basis; continued support for increased renewables to properly balance the grid; promoting a credible merit order system for power dispatch backed by a transparent auction mechanism; liberalisation of the monopolised power distribution function; and the effective synchronisation and coordination of all government policies aimed at developing the power sector will resolve the burden of excess capacity payments under the take or pay arrangements. The nemesis of successfully achieving this will be GoG policy inconsistency and that must be carefully watched.
