Reprinted with permission from the 2016 Africa Energy Yearbook
From Egypt to South Africa, Kenya to Morocco, LNG-to-power has been and is being either implemented or considered as an important component in Africa’s energy mix. Gas-fired power generation is flexible and relatively environmentally friendly, when compared to coal, so fits well as a necessary source of baseload generation to facilitate the continued growth of renewables across the continent.
Gas-fired power has been the dominant source of generation in West Africa for years due to the abundance of gas reserves. But power markets are increasingly looking to use imported gas as a fuel source, driven by the falling cost of LNG. Whilst the cost of LNG is currently reducing, LNG is not a cheap fuel source. Power from LNG tends to be less expensive than liquid fuels such as diesel (at least it is increasingly becoming de-linked from global oil prices), but it cannot yet compete with coal on cost alone in jurisdictions that do not impose a cost on carbon, meaning that governments take a short term view and back coal-fired power projects. This is being counteracted by the momentum behind carbon dioxide reduction that was achieved at COP21, even in the developing markets of Africa, together with the actions of development financing institutions and multilaterals, whose support is vital for the development of power projects in Africa.
In north Africa, Egypt has been regasifying imported LNG using its first floating storage and regasification unit since April 2015. A second FSRU was delivered in September 2015 and it now plans to charter a third FSRU to supply greater volumes of gas for power generation, due to be delivered in late 2016/early 2017. These regasification facilities will act as a bridge to the commercialisation of ENI’s supergiant 30 Tcf Zohr gas prospect. To the west, Morocco has made significant progress with its comprehensive LNG-to-power programme, issuing the request for qualification for the procurement of an onshore regasification facility at Jorf Lasfar and 2.4GW of power capacity as part of a single project, to underpin a further 4GW of power generation capacity and to supply of gas for domestic industry.
West Africa’s existing gas-fired power generation portfolio is expected to be further enhanced by a number of LNG-topower projects under development in Ghana. The case for imported LNG in Ghana is made out as a means of enhancing the stability of gas supplies from Ghana’s indigenous offshore Jubilee and Offshore Cape Three Points gas reserves. The Ghana 1000 gas-to-power project uses indigenous gas for its first 375MW phase, but then contemplates expansions up to 1300MW through the use of imported LNG.
In Southern Africa, Namibia previously selected the Xaris LNG-to-power project as the winner of a 250MW procurement process before doubts over the project surfaced. The use of LNG as a medium term bridge to domestic gas from the Kudu field would have been quite understandable for Namibia. South Africa’s significant power deficit, lack of commercialised domestic gas and proximity to international shipping routes positions it well for the development of LNG-to-power projects, which could easily use Angolan or Mozambican LNG more cost effectively than deliveries by pipeline from Mozambique. The Department of Energy of South Africa (DoE) is pushing ahead with the procurement of between 1000- 3000MW of gas-fired power generation. However, both South Africa and Namibia would need to overcome the challenges presented by having to meet significant dollar-denominated fuel costs if they were to seek to implement LNG-to-power solutions. In addition, South Africa has built up a significant track record of procuring electricity from renewable sources under a form of power purchase agreement (PPA) that may well not be sufficient to enable investors to bank LNG-to-power projects.
Fundamentally, the integration of the LNG supply, regasification and power generation components of an LNG-to-power project is very complex. The interrelationship of risks and the existence of multiple sponsor and lender groups gives rise to competing interests that make developing projects challenging at best. If the fuel supplier’s delivery obligations are guaranteed to the power generator, and the power producer’s obligation to take deliveries of gas are similarly supported, the project becomes much easier to structure. However, this rarely represents the commercial reality of power and infrastructure project development – instead sponsors want to structure the projects on a limited recourse basis and avoid significant sponsors support. Therefore lenders must analyse and get comfortable with the entire contractual matrix, including for the portion of the project that they may not actually be lending to. Clearly, the state interposing itself between various aspects of the project may significantly reduce the risks to sponsors and lenders, enhancing bankability and delivery timeframes. But that is equally challenging for economic, constitutional or regulatory reasons.
Integrated model – The purest form of LNG-to-power project would be a single entity that procures both the regasification and power infrastructure under a single financing, importing LNG for the sole purpose of producing power in a dedicated power plant. This is a somewhat high risk solution because it is the least flexible option. Any adverse performance in the power plant will directly impact the power project company’s ability to meet its gas offtake obligations and vice versa. However, this might be workable for smaller scale developments – indeed, it was the basis for the 250MW Xaris LNG-to-power project that was planned for Walvis Bay in Namibia. Developers considering this model will need to confirm that the licence conditions for power generation and gas distribution do not prevent cross-collateralisation of assets and, specifically in relation to the gas import infrastructure, that third party access rights do not exist. One interesting feature we have seen from other jurisdictions is that the LNG supply arrangements need not be firm, if supporting lenders can derive sufficient comfort with the long term availability of LNG on a spot basis, and the tariff structure under the PPA enables the power producer to pass through the gas price to the ultimate consumer or offtaker of electricity. This level of flexibility means that LNG can be procured on a spot basis, thereby avoiding significant financial commitments by the gas or power project companies, but exposing the consumer to less certain power prices.
Gas sales model
In a market with multiple consumers of gas, there will likely be more flexibility of gas offtake, making it possible to establish an LNG importation project in which the same entity imports the gas and sells it to customers. The same sponsors might separately develop an independent power project (IPP) on the basis of a long term offtake arrangement for the power (unless market fundamentals would support a merchant project), therefore providing a keystone customer for the gas project. This flexibility of gas offtake is crucial to the success of both the gas project and the power project. If construction of the power project is delayed or the power project is adversely affected by either low demand or poor performance, the gas company can sell gas to alternative customers, thereby mitigating its exposure to take-or-pay payments (usually on a cargo by cargo basis under LNG sale and purchase agreements (LNG SPAs)). Whilst this may create an impression that the power offtaker does not need to accept responsibility for take-or-pay costs under the PPA, this is rarely the case. Generally such costs would need to be passed through to the offtaker on terms that are back-to-back with the gas supply arrangements. This risk allocation can only be avoided if there is a large and relatively mature market for gas such as to give investors high levels of certainty as to the long term demand profile for LNG.
An alternative approach would be for the IPP to import LNG itself, under a long term purchase agreement with an LNG shipper, and to buy capacity rights in a regasification facility to convert the LNG to gas. This enables the regasification facility to be financed on the basis of long term stable revenues under the throughput agreement with the IPP (again, potentially developed by the same sponsors as the IPP). Excess regasification capacity can be sold to other IPPs or gas industries. This is likely to be more appropriate where the regulatory regime imposes open access requirements and therefore a portion of any regasification capacity must be reserved for the market (for example, this could be the case in South Africa in light of recent statements by the government that this is its intention). It is equally applicable where the power project is the primary driver behind both projects, as opposed to the gas importation and sale business being dominant. This is perhaps the most conventional approach to structuring LNG-to-power projects and might well be the most appropriate solution for Morocco. However, Morocco currently appears to be pursuing a fully integrated approach, which fits more closely with its historical requirement for developers to take the fuel supply risk on its other thermal power projects.
Government buyer model
Finally, a solution that might be more appropriate for a country experiencing high growth rates but with potentially uncertain long term demand for imported gas might be for the host government to act as an intermediary between LNG shippers and consumers, including IPPs. In this case the regasification infrastructure could be financed on the basis of a throughput agreement with the utility, and the IPP would purchase gas from the government, or the government could toll its gas through the power plant, whist using excess regasification capacity to stimulate local industry or domestic gas consumption. This is the approach currently being taken in Egypt, were EGAS imports LNG and sells it on to IPPs, and appears to form the basis of the Ghana 1000 project, where GNPC is expected to procure LNG imports. However, the government will need to have a relatively high degree of sophistication in order to administer the LNG SPAs, LNG trading arrangements and tolling arrangements. The diversity of domestic industry in South Africa, coupled with the potential high demand for gas, the existing role of Transnet and the conventional risk allocation under Eskom PPAs could well result in this structure being equally appropriate for South Africa, albeit that the preference of government seems to be a bundled solution.
What is project-on-project risk?
In all but one of these scenarios, the regasification and power infrastructure is financed separately and on a limited recourse basis, and in the first two cases, there is a significant interdependency between the projects, thereby creating “project-on-project” risk. This means that a default by one half of the overall project (e.g. the regasification project) will lead to an inability to generate cash flows in the other half (here, the power project). However, the limited recourse nature of both halves means there is no creditworthy entity ultimately backstopping the risk.
A good example of project-on-project risk is delays in construction of one or other component of the project. Conventionally delay liquidated damages payable under a construction contract would be calculated at a rate and with a cap so as to keep debt and equity whole for a reasonable period of foreseeable delay. This approach is satisfactory for the project to which the construction contract relates, but the same level of liquidated damages will not compensate debt and equity on, for example, a downstream power project if the delay is suffered in the upstream regasification project (or vice versa).
This characteristic of many LNG-to-power projects places a higher value on flexibility – the ability to mitigate adverse effects experienced in one element of the project on the other element of the project. In the discussion of different project structures above, one recurring theme is that the more flexibility that can be achieved at either end of the supply chain, the less risk will be assumed by the project overall and therefore the more bankable and attractive the structure will be, leading to lower power prices.
Many markets in Africa that are not accustomed to large scale gas-fired power generation have regulated tariffs that are not sufficiently flexible to either allow the power producer to pass through take-or-pay payments or adjust the price of power based on a variable underlying cost of LNG. It is perhaps important to appreciate that the price of gas may fluctuate even under long term LNG SPAs (being linked to various indices or commodity prices). Therefore the fuel charge component of the typical tariff for a gas-fired power project will need to vary to accommodate such changes. Regulators need to be comfortable with this and should ensure that the national utility or government buyer’s account is the point in the supply chain at which the cash flow impacts of faster changes in fuel costs than the regulated electricity tariff are managed.
How does technology help mitigate risk?
Small to medium scale developments offer the opportunity to utilise floating storage, regasification and unloading (FSRU) vessels as the regasification infrastructure. Capital cost can be minimised by using old vessels, converted for this purpose and the current levels of competition in the vessel chartering market mean that financing new builds or conversions is relatively straightforward and flexible (in terms of the structural requirements of vessel financing lenders). FSRUs are chartered by the entity operating the regasification portion of the project and are generally financed separately from the other infrastructure. Using a floating regasification solution offers an LNG-to-power project the flexibility to mitigate (in part) construction delays or major equipment breakdown during operations of the power project by delaying the FSRU’s delivery date (within reason) or even through redeployment of the vessel as an LNG carrier to generate extra revenue pending completion of the IPP.
Technology can also assist power projects to overcome the risk of non-availability of gas. Most CCGT plants are capable of being configured to operate on liquid fuel as a back-up, when natural gas is unavailable. This flexibility is specifically recognised in the DoE’s request for information in relation to early power generation, which helps to mitigate gas availability risk during the initial operations of the gas import facilities. However, operating on liquid fuel has adverse effects on plant performance and increases both operation and maintenance costs to varying degrees, depending on the gas turbine technology. Provided this risk is addressed adequately under the PPA, the use of a dual-fuelled plant offers the IPP the ability to mitigate its exposure to delays in completion of the regasification infrastructure and to short term LNG unavailability (throughout the operating period).
Demand risk – A 20-year take-or-pay contract for LNG is a significant financial commitment. It will generally require the buyer to lift LNG volumes in accordance with a fairly restrictive schedule (essentially imposing a take-or-pay obligation on the buyer). Dispatch risk almost always lies with the offtaker in any gas-fired IPP in an emerging market (insofar as capacity charges will be payable for availability, regardless of power demand). However, the concept of the national utility being responsible for take-or-pay payments arising from the LNG supply chain may be uncomfortable for African power purchasers, but this is common practice in other markets around the world and is an important component of the bankability of LNG-to-power projects. Utilities are responsible for capacity planning and demand forecasting, so to seek to place responsibility for take-or-pay payments with the generator or gas supplier would be contrary to the principle of ensuring that the risk lies with the person best able to manage it. By procuring some or all of their LNG requirements on a short term basis, utilities can make a trade-off between cost and committed volumes that may be desirable and consistent with economic purchase obligations.
Conversely, take-or-pay risk can be mitigated to some extent by having multiple offtakers of gas. This is particularly helpful if the volumes to be delivered by the gas project company to other customers are flexible. It may also be possible for the gas supplier to sell excess LNG cargos on the spot market, thereby mitigating its exposure to take-or-pay payments. This flexibility is subject to redirection rights under the LNG SPA, which are becoming increasingly flexible, thereby enhancing international short-term trade in LNG. There is currently no formal market price for spot cargos of LNG, so it will be challenging for a developer to fully mitigate, and therefore assume, this risk at the outset of a project. An independent developer in control of the full value chain will, however, more likely be able to manage these risks in an efficient manner.