In 2014, there was a marked upturn in interest in solar projects in the African market. Some project finance lenders have not financed solar PV previously, or not in sub-Saharan Africa, and are taking a cautious approach; others are increasingly comfortable with solar PV as an asset class. This note, first published in June 2015, outlines financing issues encountered in solar PV projects in sub-Saharan Africa, particularly South Africa, and across the continent.
Solar energy has most definitely arrived in Africa. CSP plant projects have completed in South Africa (Khi and KaXu) and Morocco (Noor I). There are solar PV projects seeking finance in Nigeria, Ghana, Senegal, Kenya, Namibia and Zambia; in Rwanda and South Africa, projects have already reached financial close.
In many jurisdictions, the term of the power purchase agreement (PPA) is from 20 to 25 years, with a tariff structured to match this period. A trend towards longer tenors for PPAs is being driven by a combination of host governments seeking a lower tariff, outside South Africa, and looking to match the tenor with the anticipated useful life of the solar PV plant.
Traditionally, it has been difficult to achieve debt tenors to match a PPA term of 20 to 25 years on African power projects; a debt tenor of even 15 years has been a stretch. Development finance institutions (DFIs) generally have more capacity to provide long-term debt, particularly compared to commercial banks, though debt tenors offered by DFIs are still unlikely to cover the full term of the PPA.
Recently, a few lenders have started accepting tenors of 17 to 18 years plus the construction period.
As an intermittent energy source dependent on irradiation levels, the financial model base case for solar PV depends on the forecast of electricity that will be generated by the power plant over the term of the loan. Forecast generation is calculated using ‘Probability’ or P values: a P90 forecast is a more conservative forecast (representing a 90 per cent probability that the electricity generated will exceed this value) and P50 is the average (a 50/50 chance that the electricity generated in the period will exceed or fall below this value).
There is no long-term track record for the performance of solar PV power plants in Africa. However, the expectation is that the solar PV projects will perform well in most African environments, especially as the currently installed global solar PV fleet is generally performing in line with P50 forecasts (out-performing the more conservative forecasts).
In the first round of the South African REIPP programme, lenders tended to insist on P90 forecasts for the base case in the financial model. This contrasts with developed markets, such as Europe and the United States, where P50 is usually adopted. We are now starting to see a move in South Africa towards P50 forecasts being used for the base case.
Most of the DFIs and multilateral finance institutions (IFIs) will probably look to use P90 levels until it is proven that solar PV projects in Africa have performed in line with P50 or P75 expectations for several years.
In African solar PV project financings, the debt service cover ratios accepted by lenders have been in the range of between 1.25 and 1.35. The energy forecasts underpinning such DSCRs have usually been on a P90 basis.
In mature markets like Italy, the norm has been a DSCR of 1.3 using a P50 forecast.
In reverse auction processes, such as the REIPP programme in South Africa, we are starting to see more pressure on the base case ratios as sponsors and lenders seek to optimise the tariff.
Unlike thermal power projects, most solar PV projects will have a fixed tariff for the term of the PPA and predictable operating characteristics and costs. Lenders should therefore be able to rely on historic ratio tests only. Forecast DSCRs and loan life cover ratios (LLCRs) are most useful when the project's revenues and/or load factors are variable, or where costs are expected to fluctuate significantly over time or year to year.
Lenders tend to require compliance with forecast ratios for distributions and, in some cases, to avoid triggering a default. Once lenders are reassured that solar PV projects around Africa are performing solidly and in line with forecast expectations, they may start to relax these requirements.
In many jurisdictions in sub-Saharan Africa, the ratio between debt to equity on a power project has been traditionally close to 70:30. There has been no debt-to-equity ratio in excess of 80:20. Developers have to strike a balance between the low tariffs required by host governments and an economically attractive investment rate of return (IRR) for equity investors. This is particularly so given the perceived risk for equity investors of investing in Africa power projects.
Faced with host governments encountering pressure from regulatory authorities and consumers to keep PPA tariffs as low as possible, and with debt tenors unlikely to match the tenor of the PPA, sponsors are seeking to increase the gearing ratio of project companies beyond traditional levels.
Some lenders require sponsors to inject equity upfront; this usually reflects the credit quality of the sponsors and does not relate to any specific African or solar PV risk issues.
Some DFIs and IFIs insist on an ongoing debt-to-equity ratio test which must be met during the life of the project. This is not a customary project finance requirement, as project finance traditionally looks at revenue expected to be generated by the project, and not the balance sheet of the project company. The impact of this ongoing test is that sponsors cannot prepay shareholder loans any faster than is envisaged in the base case, otherwise they will breach the debt-to-equity ratio. This can be even more problematic when the base case is determined using P90 levels as, where the power plant generates at P50 levels, this can result in cash being trapped in the project company.
Outside South Africa, most solar PV projects are financed in US dollars (or, less commonly, in euros). This reflects the tariffs which are generally either denominated in US dollars or in local currency but are either linked to or paid in US dollars.
Local currency financing is available in some markets but tenors are generally short and pricing is much higher than dollar- or euro-denominated debt.
Margins vary significantly from project to project based on the usual dynamics of country risk, competitive tension between lenders, sponsor relationships and prevailing market conditions.
There is downwards pressure on margins in many African jurisdictions. This reflects in part the improved macro-economic conditions but also the increasing number of attractive solar PV projects which has resulted in funders ‘crowding in’ on bankable projects.
To put African solar PV in context, in recent months the margins for European project finance have fallen significantly, reflecting reduced country risk and improved liquidity. It is not uncommon for sponsors to achieve margins below 200 basis points for European PV projects. There are signs of pricing pressure in South East Asia, Latin America and other markets. African pricing has not always followed other markets but the increase in deal flow across the continent should help reduce margins in coming years.
It is usual to expect sponsors to have a debt service reserve account (DSRA) funded by no later than completion of the project with an amount equal to six months of debt service payments.
Occasionally, lenders may also require sponsors to set up a maintenance reserve account (MRA), which will cover anticipated major maintenance expenses. Whether an MRA is needed depends on the contractual structure of the operations and maintenance arrangements for the project and the underlying warranties on the modules, inverters and other key components of a solar PV power plant, such as the frames or trackers.
A notable example of this is inverter replacement, which often occurs approximately 7 to 10 years after the commercial operations date. In some cases lenders will insist on an inverter reserve account, which requires the project company to set aside funds on a six-monthly basis to meet the anticipated costs of replacing inverters.
Lenders financing a well-structured solar PV project should not require completion support from the sponsors.
It is customary (and reasonable) for lenders to expect sponsors to have a sensible level of contingency set aside - taking into account the location and design of the project and the counterparties – but this is a question of project risk in general rather than any specific risk relating to PV. In Africa, the level of contingency required by the lenders may be higher than other regions due to concerns on issues such as change-in-law risk and the limited number of projects developed to date.
However, the concept of a full or partial completion guarantee from sponsors, in addition to a robust construction contract, can be resisted, unless there are other factors that affect the bankability of the project.
Location should not make a difference as to whether completion support is required. It is not customary for lenders to require completion support from sponsors on thermal projects in Africa over and above the contingencies (which may be provided by way of standby equity) and solar PV need be no different.
The story may be different if the project utilises new technology. Solar PV cannot, for example, be compared to CSP (concentrating solar power), a less-established technology which commonly requires a certain level of completion support.
In project finance a ‘cash sweep’ is often used as means of prepaying debt if the performance of the project (and therefore available cash) has exceeded a given level or if the project has under-performed and lenders are keen to reduce debt quickly.
Cash sweeps are not a prerequisite to financing solar PV projects. That said, they remain part of the tool kit for many lenders; cash sweeps are commonly found in project finance transactions where the lender is keen to shorten the nameplate debt tenor.
On some solar PV projects in Africa lenders have looked for a downside cash sweep – where failure to meet lock-up levels results in an immediate or potential cash sweep. This is often a request of DFIs.
Distribution conditions represent an area of some contention between sponsors and lenders. If there are credit concerns with a project, lenders and advisers may seek to extend the list of conditions that must be met prior to equity distributions being made to sponsors. The aim of this is to lock in the project's cash as an additional buffer in case of hard times in the future.
There are lock-up conditions which may be applicable to solar PV. These can be categorised as:
- a certain requirement of lenders
- a reasonable requirement of lenders, depending on circumstances
- generally to be resisted by sponsors.
- historic DSCR above lock-up level
- no default
(this is default, not just event of default, as it is not reasonable to expect a distribution where a default has occurred and the occurrence of an event of default is subject only to a grace period)
- reserve accounts fully funded
- first repayment made
- completion achieved.
- forecast DSCR/LLCR above lock-up levels
- no outstanding disputes under construction contracts which are not fully covered by cash reserves/contingencies
- no material grid failures
- no material deterioration in offtaker creditworthiness
- minimum level of working capital to be retained within the project company
- no change of law which has a material adverse impact
- no unavailability of material insurance.
- lock-up levels which are the same as the base case or debt sizing ratios
- no distributions for first two years of operation
(to give lenders additional comfort in case the project faces teething problems at the start of operations)
- any breach of the PPA, whether or not material.
With the entry of proven, creditworthy top-tier sponsors into the African power market, equity bridge facilities are coming into play. These are used to reduce the tariff bid in an auction process, improving the equity IRR for the sponsor.
The short construction period for solar PV projects (months compared to years for other power projects) means that sponsors only receive the benefit of delaying their equity contribution for a short time. The costs of arranging and documenting an equity bridge facility may therefore outweigh the benefit to sponsors.