A 220-bed private hospital in Lagos lost its duty 500 kVA set on a Friday evening because a fuel filter had been left in service for 900 hours on a 500-hour interval. The standby set picked up the load, ran for eleven hours on a partially blocked strainer, and dropped out at 02:40. Theatre and cold-chain were on inverters that were never sized for that duration. The maintenance contract in the drawer said “quarterly servicing” and nothing else, so nobody was in breach.
That is the real problem with most of what passes for a generator maintenance contract in Nigeria. The document is priced, signed and filed, but it does not define what work is done, at what interval, with what parts, within what response time, or what happens when the set is unavailable. This article sets out the clauses a technical buyer should insist on before signing, and what each tier realistically costs.
What a Generator Maintenance Contract in Nigeria Should Actually Cover
A maintenance agreement is a risk transfer instrument, not a servicing subscription. The only question worth asking during evaluation is: after signature, which failures are the contractor’s problem and which are still yours?
A complete agreement addresses eight things. Asset schedule with serial numbers, ratings and duty classification. Scope of work tied to running hours. Parts and consumables responsibility with named brands or approved equivalents. Response and rectification times by fault severity. Spares holding and lead time commitments. Testing and reporting obligations, including load bank testing. Exclusions stated in plain language. Remedies when the contractor misses the terms.
Anything missing from that list becomes a variation order later, usually at the worst moment.
Duty classification is the clause most buyers skip. A set rated for standby duty under ISO 8528 is not warranted for continuous running, and a contractor who services a standby-rated machine that you actually run 4,000 hours a year is servicing it against the wrong interval. Get the rating, the annual running hours and the average load factor written into the schedule. If you are unsure of your load factor, that is a measurement exercise before it is a contracting exercise.
If you are re-tendering an existing arrangement, request a technical proposal with the asset schedule attached so that quotes are comparable line by line rather than lump sum against lump sum.
Scope of Work: Running Hours, Not Calendar Months
Diesel engines wear by hours and load, not by quarters. A contract that promises “four visits per year” against a set running 14 hours a day is arithmetic that fails on contact with the plant.
Write the scope as an hours-based matrix. Typical intervals for a mid-range industrial diesel set, subject to the OEM manual and site conditions, run as follows.
- Daily or per-start checks by your own operator: oil and coolant level, fuel level, leaks, battery condition, panel alarms.
- 250 hours: visual inspection, belt tension, battery electrolyte and charging voltage, air filter restriction check, alarm test.
- 500 hours: oil and oil filter change, fuel filters, air filter as indicated, coolant test, valve clearance check on schedule.
- 1,000 to 1,500 hours: injector inspection, cooling system service, alternator bearing check, insulation resistance test, protection relay verification.
- 6,000 to 8,000 hours: top-end work depending on oil analysis trend and blow-by.
- 12,000 to 20,000 hours: major overhaul, planned against an outage window, not attempted as a breakdown response.
Two conditions in Nigeria compress these intervals. Dust loading during harmattan shortens air filter life sharply on unsheltered installations, and fuel quality varies enough that filter change intervals should be driven by differential pressure and by what comes out of the tank, not by the calendar. Insist that the contract allows interval adjustment on evidence, with a stated mechanism for pricing the extra visits.
Oil sampling is the cheapest clause in the document and the one that pays for itself first. Quarterly sampling with wear metal, viscosity, soot and coolant contamination analysis will flag a liner or bearing problem months before it becomes an engine-out event. Make the lab report a deliverable, not a favour.
Response Times, Escalation and Penalties That Bite
“Prompt response” is not a term. A working agreement defines severity levels and puts a clock on each one.
A defensible structure for a single-site industrial user in Lagos, Port Harcourt or Abuja looks like this. Severity 1, total loss of power with no standby available: telephone response within 30 minutes, engineer on site within 4 hours, continuous work until restoration or a written interim plan. Severity 2, duty set down with standby carrying load: engineer on site within 12 hours. Severity 3, degraded performance or alarm without loss of load: attendance at the next scheduled visit or within 5 working days.
Response time and rectification time are different obligations. A contractor who arrives in 4 hours and then waits three weeks for a part has met the first and failed the plant. Specify both, and specify what happens when a part is genuinely unavailable, which in this market is a real possibility rather than an excuse.
Travel distance changes what is honest. A four-hour on-site commitment in Kano for a contractor based in Lagos is not credible unless there is a resident engineer or a formal partner arrangement, and you should ask which it is. Coverage across Lagos, Rivers, FCT, Kaduna, Delta and Kano is a mobilisation question, and the answer belongs in the contract as named response bases.
Penalties should be proportionate and enforceable. Service credits against the monthly fee, capped at a percentage of the annual value, are workable. Consequential loss claims are not, and a contractor who accepts unlimited liability for your production loss is either misreading the clause or not intending to honour it. Base the credit on measured availability, and define how availability is measured before the first invoice.
Spares, Consumables and Lead Times in the Nigerian Market
This is where a generator O&M agreement is won or lost. Filters and lubricants are usually available in Lagos within days. Injectors, turbochargers, AVRs, ECUs and alternator windings for less common engine families are not.
A 750 kVA prime-rated set at a rice mill in the North West lost output over three weeks and was eventually traced to four failed injectors. The injector set landed at roughly NGN 3.6m (about USD 2,400 at NGN 1,500 to the dollar, so verify the rate on the day), but the damage was the seven-week lead time. The mill hired a 500 kVA rental at NGN 2.1m per month plus fuel and ran at reduced throughput. A clause requiring the contractor to hold a critical spares list on site would have cost a fraction of that.
Build a critical spares schedule into the agreement and name who owns the stock. For most industrial diesel sets that list includes a full filter set, one AVR, one set of injectors, water pump, thermostat, belts, sensors, a starter motor or its brushes, and the batteries. Where a set is single and unbacked, the list should extend to a turbocharger cartridge.
Ask for the contractor’s parts sourcing route in writing. OEM channel, authorised distributor or open market are three different risk profiles for price, warranty and counterfeit exposure. The same logic applies further up the plant, and the reasoning is set out in more detail in our note on turbine spare parts lead time.
Also settle the equivalents question. “OEM or approved equivalent” without a named approver means the contractor decides, and you find out what was fitted during the next failure investigation.
Comparing O&M Contract Tiers: Price Against Risk Retained
The table below gives indicative annual figures for a single 500 kVA prime-rated diesel set running roughly 4,000 hours a year at a site within the contractor’s base state. Fuel, major overhaul, rental cover and travel outside the base state are excluded throughout.
| Contract tier | What the contractor delivers | Who buys parts | Indicative annual cost | Risk you retain |
|---|---|---|---|---|
| Inspection and reporting | Scheduled inspections, oil sampling, written condition report, no corrective work | You | NGN 1.8m to 2.6m (USD 1,200 to 1,700) | All labour and all parts on failure |
| Preventive maintenance | Hours-based servicing, labour and routine consumables, alarm and protection testing | You buy non-routine parts | NGN 5.4m to 7.6m (USD 3,600 to 5,100) | Corrective repairs, major components |
| Comprehensive O&M | Preventive plus corrective labour and parts up to a stated cap, annual load bank test | Contractor to cap | NGN 11m to 16m (USD 7,300 to 10,700) | Above-cap components, overhaul, fuel system contamination |
| Comprehensive with availability guarantee | All of the above plus on-site critical spares, defined availability target and service credits | Contractor | NGN 18m to 28m (USD 12,000 to 18,700) | Fuel supply, operator error, civil and installation defects |
Two observations from tendering exercises. The gap between tier two and tier three is usually smaller than buyers expect once a realistic parts budget is added to the cheaper option. And the top tier only makes commercial sense where an hour of downtime costs more than the annual premium difference, which is worth calculating rather than assuming. Our breakdown of plant downtime cost per hour sets out a method for that.
Pricing a Generator Servicing Contract Against Your Downtime Cost
A cold store in Port Harcourt running a single 250 kVA set treated maintenance as an overhead to be minimised, at roughly NGN 3.1m a year. One extended outage during a grid failure, six hours in a chamber holding fish and poultry, wrote off about NGN 4.1m (USD 2,700) of product in a single afternoon, before the cost of the repair itself.
The calculation is straightforward. Estimate your loss per hour of unplanned outage, including product, labour standing idle, contractual penalties and restart losses. Multiply by a realistic expected outage hours figure for your current arrangement. Compare against the premium for the next tier up. If the arithmetic is close, the deciding factor is usually how long a critical part takes to reach your site, not the service schedule.
Load bank testing deserves its own line item. Standby sets that spend their lives at 20 to 30 percent load will glaze bores and wet stack, and the first time anyone discovers the set cannot hold full load is the day it has to. An annual four-hour test at stepped load on a 500 kVA set runs around NGN 1.1m to 1.4m including the bank and cabling, and it is the only honest proof of capacity. The detail is covered in our guide to generator load bank testing.
Watch for four red flags in a draft agreement. A fixed number of visits with no hours reference. Parts described as “as required” with no cap and no price list. No named response bases for your sites. And an availability guarantee with no definition of how availability is calculated.
Two regulatory points are worth confirming for your own installation. Self-generation above 1 MW requires a captive generation permit from the Nigerian Electricity Regulatory Commission, and installations are subject to inspection and certification regimes administered by the Nigerian Electricity Management Services Agency. Neither obligation transfers to a maintenance contractor by default, so state in the agreement who prepares and holds the documentation.
Where a site has poor records, undocumented modifications or a history of repeat failures, the sensible first step is a plant assessment rather than a contract. A structured power plant audit establishes actual condition, load profile and remaining life, which is also what makes competing bids comparable.
Getting to a Contract You Can Enforce
Bring three things to the table before you invite bids: an accurate asset schedule, twelve months of running hours and fuel consumption, and an honest downtime cost figure. With those in hand, tier selection becomes an arithmetic exercise rather than a negotiation over adjectives.
Our engineers can review an existing agreement clause by clause against the schedule above, or scope a new one from a site visit. See generator maintenance in Nigeria for the service scope, and generator and turbine maintenance cost for how the numbers are built.
To move forward, book a plant assessment or call +234 803 000 0000 to discuss coverage for your sites.
Frequently Asked Questions
How much should a generator maintenance contract cost in Nigeria?
For a single 500 kVA prime-rated set at around 4,000 running hours a year, preventive maintenance typically falls between NGN 5.4m and 7.6m annually, while comprehensive O&M with a parts cap sits between NGN 11m and 16m. Fuel, major overhaul and rental cover are normally excluded. The figure moves with engine family, site access, running hours and the exchange rate, so any quote should be built from your asset schedule rather than a rate card.
Should the contract be based on calendar visits or running hours?
Running hours, with a calendar backstop for sets that run very little. Engine wear tracks hours and load factor, so a quarterly visit schedule under-services a hard-worked set and over-services a rarely used standby machine. A backstop of one visit every six months protects lightly used sets against battery, coolant and fuel degradation, which are time-driven rather than hours-driven.
What should the response time clause say?
It should define severity levels, and give each one a telephone response time, an on-site attendance time and a rectification target. Total loss of power with no standby available should carry the tightest commitment, typically 30 minutes by phone and four hours on site within the contractor’s base state. Attendance times for distant states should reflect where the contractor actually has engineers, and the contract should name those bases.
Does a maintenance contract cover a major overhaul?
Usually not, and you should assume it does not unless the document says otherwise in a numbered clause. Overhauls at 12,000 to 20,000 hours require an outage window, workshop facilities, specialist measurement and qualified engineers on site, and they are priced as separate projects. A good agreement instead commits the contractor to forecast the overhaul from oil analysis and performance trending so it can be budgeted and planned rather than triggered by a failure.