Categories
Mining Infrastructure

Standby generators keep a mining site running through an outage, but only if someone maintains them. How to decide between a service contract and your own crew.

A standby generator earns its keep on the days nobody thinks about it. It sits behind the mining containers, gets exercised once a month, and is forgotten again until the grid actually fails. That gap between purchase and the one moment it matters is where maintenance quietly decides whether the investment pays off or becomes an expensive paperweight.

Sizing and specifying the unit is an engineering decision, covered well elsewhere. Keeping it ready to start on the worst day of the year is a different kind of decision, and it is a commercial one: do you put the generator under a service contract with the manufacturer or a local provider, or do you build the capability in-house. Most hosted and self-hosted operators never frame it as a choice. They default to whichever option the installer mentioned first, and live with the consequences for the next ten years.

What a service contract actually buys

A maintenance contract is not really buying labor hours. It is buying three things that are hard to build yourself: parts access, documented history, and a defined response time.

Parts access. OEM and authorized-dealer contracts typically guarantee stock on wear items – filters, belts, injectors, coolant, the sensors that fail first – and a supply chain for the parts that do not sit on a shelf. A crankshaft position sensor or a control-board replacement can have a lead time measured in weeks if you are sourcing it cold during an outage. Under a contract, that part is often already staged regionally.

Documented history. Every oil analysis, load-bank test and fuel-quality check becomes a paper trail. That history is what a warranty claim depends on, what an insurer wants to see after a failure, and what a buyer wants to see if the facility is ever sold. A generator maintained inconsistently by whoever was available that week has no comparable record, even if the actual work was fine.

Defined response time. Contracts specify a response window – often 4 to 24 hours depending on tier and region – and that number is contractual, not best-effort. For a site running at the capacity numbers covered in power redundancy engineering, where the entire point of the generator is covering the gap between a grid failure and resumed production, a response time that is merely “whoever is free” undermines the redundancy tier you paid to build.

The cost is real and ongoing: annual service agreements commonly run a meaningful percentage of the generator’s replacement value, scaling with run hours and site count, and the invoice arrives whether or not the unit ever actually starts on battery power that year.

What self-maintaining actually buys

The case for an in-house crew is not usually cost on paper – a part-time technician plus parts inventory can look cheaper than a contract, but rarely is once overhead is counted honestly. The real case is control and speed for sites that already run their own electrical staff.

If a facility already employs technicians for switchgear, breakers and the electrical infrastructure described in transformer and switchgear maintenance, extending that team to cover the generator is often a marginal addition rather than a new discipline. The same crew that already walks the electrical room weekly can run the monthly exercise cycle, catch a failing battery before it strands the unit, and respond at 2 a.m. without waiting on a dispatcher.

Self-maintaining also removes a scheduling dependency. A contracted technician serving a region covers multiple sites and multiple customers; a demand spike after a regional storm can push a 24-hour response window to 48 in practice, even when the contract says otherwise. An in-house crew answers to one site.

The honest cost of self-maintaining is harder to see upfront: training, a parts inventory that someone has to fund and track, and the institutional risk of a single technician who knows the unit leaving the company. None of those show up on a quote the way a service contract’s annual fee does, which is exactly why they get underestimated.

The decision in practice

Four factors tend to settle it more reliably than a cost comparison alone.

Generator count and site count. One generator at one site rarely justifies a dedicated in-house maintenance capability; the fixed cost of training and parts inventory does not spread across enough equipment. A multi-site operator running several units under the kind of sizing discussed in generator sizing and selection starts to reach the scale where an in-house team pays for itself.

Existing electrical staffing. If the site already has electricians on shift for other reasons, extending their scope costs less than it looks like it should. If the site runs lean with no electrical staff at all, bringing in a maintenance capability purely for the generator is usually the more expensive path.

Warranty terms. Many manufacturer warranties on larger standby units are conditioned on maintenance performed to a documented schedule, sometimes specifically by an authorized provider. Read the warranty before assuming self-maintaining is free of consequence – a lapsed warranty on a unit that fails in year three can cost far more than any service contract would have.

How much the generator actually has to do. A unit backing up a facility built on the redundancy tiers above N+1, or paralleled with others under multi-genset synchronization, carries more operational weight than a single standby unit at a small site. The more the architecture depends on the generator actually starting correctly, the more a documented, contractual response time is worth paying for.

The hybrid most facilities land on

Few operators run a pure version of either approach once the fleet has been through a few seasons. The common middle ground splits the work: an in-house crew handles the monthly exercise cycle, visual inspection and basic fluid checks, while a contracted service covers the annual load-bank test, major component work and anything touching the warranty. This keeps routine attention cheap and frequent while preserving the parts pipeline and documented history a contract provides for the work that actually matters when something fails.

That split also solves the single-point-of-failure problem on the people side. If the in-house technician who knows the generator leaves, the contracted provider still has the maintenance history and can step in without starting from a blank record. A generator maintained entirely by one person’s institutional memory, with no contract and no paper trail, is a fragile arrangement no matter how competent that person is.

Questions worth asking before signing a service contract

Not all service contracts are equivalent, and the differences rarely show up until the one time the generator is actually needed. A few questions separate a contract worth paying for from a logo on an invoice.

Is the response time actually contractual, or a target? A “target response time” in marketing material and a “guaranteed response time” in the service-level section of the agreement are not the same commitment, and only one of them has a remedy attached when it is missed.

Who performs the work? Some contracts are sold by one company and subcontracted to whoever is regionally available, which can mean a technician unfamiliar with the specific unit showing up on the day it matters. Ask whether the technician assigned to your site is dedicated, rotating, or subcontracted, and whether you get continuity of the same person or crew across visits.

What does the annual fee actually include? Time-and-materials contracts bill routine visits at a low rate and then bill parts and major repairs separately, which can look cheap until the first real failure. Fixed-fee contracts bundle more of that risk into the annual number. Neither structure is wrong, but comparing a time-and-materials quote against a fixed-fee quote on headline price alone compares two different products.

How often is the load-bank test actually performed, and who sees the results? A generator that starts and idles during a monthly exercise can still fail to carry full load when it counts. An annual load-bank test under real load is the check that catches that failure mode before an actual outage does, and the report from it should go to you, not stay filed with the service provider.

What this is not a substitute for

Maintenance sourcing does not fix a generator that was undersized for the load it is expected to carry, and it does not fix a synchronization scheme that was never load-tested under real paralleling conditions. Those are the engineering decisions covered in the sizing and paralleling guides above, and no maintenance contract compensates for getting them wrong at the design stage. What maintenance sourcing decides is whether a correctly specified unit is actually ready when the grid goes down – which, for a piece of equipment whose entire job is working on the one day it is asked to, is most of the value it was bought for.

The equipment-side analog is the ASIC fleet itself: the same build-versus-buy logic that applies to preventive maintenance on hash boards and cooling applies to the generators that keep those ASICs running through an outage. Neither decision is about whether maintenance happens. It is about who is accountable for it happening on schedule.

If you are weighing a service contract against building in-house capability for a mining site’s backup power, get in touch.

Explore Rax Mining

Categories