The file arrives with a site, twelve months of resource measurement and a price per kilowatt-hour that stands comparison. Three months later the credit committee has sent back a list of questions, and none of them is about solar. They are about a lease, a letter from the grid operator, the termination clause of a contract. The mismatch is not an anomaly. It is the appraisal doing its job.
A project lender, whether a commercial bank, an international financial institution or a concessional fund, does not finance a plant. It finances a revenue stream over fifteen to twenty years, with no or limited recourse to the shareholder. Its question fits in one sentence: who pays me, with what certainty, and what happens if it stops. Everything else in the appraisal follows from that sentence.
The offtake contract, and above all the offtaker
In Morocco a solar project sells its output in one of two ways. Either to the national electricity and water utility, ONEE, following a tender or a negotiated contract. Or to an end customer, industrial or public, under the framework opened by Law 13-09 and its successive amendments, which extended grid access down to medium voltage. The self-generation regime of Law 82-21 adds a third route for consumers who produce for their own use.
In the first case the lender looks at the public counterparty and the guarantees around it. In the second it looks at the customer’s financial strength, the length of the contract against the length of the debt, and what happens if the customer closes the plant. A ten-year contract does not finance an eighteen-year loan. A customer with no credit rating is replaced by a bank guarantee, a reserve account or a substitution mechanism, and each has a price.
The clauses read first: take-or-pay or deemed-energy payments, tariff indexation, the currency of payment and of debt, force majeure, termination compensation. If the tariff is in dirhams and the debt in euros, someone carries the exchange risk. The committee wants to know who, and at what cost.
Land
This is the slowest item, and the one people believe is settled earliest. A letter of intent from a commune is not a title. A promise of a lease over collective land is not a lease. The lender asks for a real right or a lease that outlasts the debt, a title free of dispute, and evidence that the compensation provided for has actually been paid to those entitled to it.
Morocco’s large projects showed how it is done: at Ouarzazate the land was acquired and secured before investors were consulted, which removed from the project a risk none of them would have agreed to carry. A developer without that capacity must at least produce the full chain: legal status of the land, applicable procedure, competent authority, a realistic timetable, and the compensation budget inside the financing plan rather than in a footnote.
Grid connection
A letter saying the connection “raises no particular difficulty” carries no weight with a credit committee. What counts is a connection study carried out by the grid operator, with the available capacity at the target substation, the route and cost of the line, how that cost is shared, a committed date, and the conditions under which the operator may limit injection. Projects that lose two years almost always lose them here, because the queue at the substation was unknown when the feasibility study was written.
One point is often overlooked: curtailment risk. A plant that generates but cannot inject does not invoice. If the contract does not pay for energy not taken for grid reasons, the financial model must allow for it, and the debt is sized on what the plant will sell, not on what it will produce.
The permitting sequence
The authorisation under Law 13-09, environmental acceptability under Law 49-17 on environmental assessment, building permits, the opinions of the administrations concerned: each has its lead time and, above all, its conditions. The lender does not only ask for the list. It asks for the sequence. Which permit conditions which. Which document is missing on the day of the committee, and what its issue depends on. A long-stop date in an offtake contract that leaves no time to obtain the last permit is a default before commissioning.
The sponsor
Equity must be real and committed, not promised subject to a third party signing. Operating capability must be demonstrated, through the sponsor’s own record or a contract with a recognised operator. The resource study must be bankable: at least one year of on-site measurement, reconciled with satellite data, and output expressed as probabilities, the median value for expected yield, the ninety-per-cent value for sizing the debt. A financial model whose debt service cover falls below the threshold in a single sensitivity case gets fixed before the committee, not after.
What is displayed, and rarely decisive
Technology. Modules, inverters and structures are catalogue products, warranted by manufacturers lenders already know. The technical choice gets checked; it is almost never debated.
Tariff. The lowest tariff attached to an unbankable contract loses to an average tariff attached to a solid one. Committees do not compare prices. They compare probabilities of repayment.
The environmental and social package. The IFC Performance Standards and the equivalent frameworks of other lenders are a condition of entry. A file that meets them gains nothing; a file that ignores them is set aside unread. The difference is made by the quality of engagement with local stakeholders, because that is what protects the construction schedule.
In what order to work
The useful order is the reverse of the appealing one. Start with what is slow and irreversible: land, grid connection, the counterparty and its contract. Finish with what can be bought: equipment. In between, build the permitting sequence and size the debt on prudent revenue.
A well-prepared file is one where every question the committee will ask already has its attachment. This is not a matter of volume: the three-hundred-page files that fail are missing the grid operator’s letter. It is a matter of order.