Most backup power conversations at mining sites stop at the generator. The nameplate gets checked, the transfer switch gets specified, the load bank test gets scheduled, and everyone moves on. The question that decides whether any of that matters on the day is simpler and far less discussed: how long can it actually run, and will the fuel still be good when you ask it to.

This is about the fuel side of backup power. It is a different problem from containment, which is about what happens when fuel escapes. This is about supply continuity, which is about having enough and having it usable.

Run-time is a calculation, not a feeling

The number that matters is not tank capacity. It is hours at your actual load.

Generator fuel consumption varies with load factor, and it is not linear in the way people assume. A unit running at a quarter of its rating does not burn a quarter of the fuel it would at full rating. Consumption curves are published by manufacturers for each load step, and that curve is what you size against.

Start from the load you will genuinely carry during an outage, which at most mining sites is not the full site. It is the critical subset: controls, network, cooling for whatever is still running, lighting, and the safety systems. If your plan is to curtail the miners and keep the building alive, the backup load might be a small fraction of site demand, and the run-time arithmetic changes completely in your favor.

Work out the hours you need, not the hours you would like. That figure comes from the realistic restoration time for your area and the time it takes to get a fuel truck to you. The second number is frequently longer than the first, and it is the one people forget.

Usable capacity is less than tank capacity

Three deductions sit between the number on the tank and the fuel you can actually burn.

Fill limits. Tanks are not filled to the brim. A proportion of the volume is left as expansion space, and the practical fill level is below nominal capacity.

Unusable bottom volume. The suction point sits above the tank floor, deliberately, so that accumulated water and sediment are not drawn into the engine. Everything below that point is not available to you.

The reserve you should not touch. Running a tank to the bottom pulls exactly the material the suction height was designed to avoid. Treat the last portion as unavailable and size as if it does not exist.

Between them, these mean planning against nominal capacity overstates your real run-time. The correction is not complicated, it just has to be applied rather than assumed away.

Stored fuel degrades, and diesel degrades faster than people expect

This is the part that catches sites which did everything else right.

Diesel is not indefinitely stable. Left sitting, it oxidizes and forms sediment and gums that block filters. Modern low-sulfur formulations and biodiesel blends are generally less stable in storage than older fuel, which means practice learned a decade ago may no longer apply to what is in your tank today.

Water is the bigger problem, and it arrives even in a sealed tank. Air drawn in as the tank breathes carries moisture, which condenses on cool internal surfaces and collects at the bottom. Water supports microbial growth at the fuel and water interface, and that growth produces the sludge that plugs filters, typically at the worst moment, because that moment is when the engine is finally being asked to run hard.

The practical regime is unglamorous and it works:

• Test rather than assume. Periodic fuel sampling tells you about water content, microbial activity and stability. It is the same discipline as oil analysis on the engine itself, applied one step upstream.
• Drain water from the low point on a schedule, because that is where it goes and it will not leave on its own.
• Treat fuel that will sit. Stabilizers and biocides exist for exactly this case. Biocide is a response to a confirmed problem rather than a routine additive, so test first.
• Polish or turn over the stock. Fuel polishing filters and circulates the stored volume. Where the site has another use for diesel, rotating stock is simpler and cheaper than polishing it.
• Keep the tank as full as the regime allows. Less headspace means less air, which means less moisture drawn in.

Gas-fueled sites have a different exposure

A site running on pipeline or wellhead gas does not store fuel, which removes the degradation problem entirely and replaces it with a dependency problem.

With no on-site inventory, an interruption upstream is immediate. There is no buffer. Whether that matters depends on the firmness of your supply arrangement and on what happens to that arrangement during a regional demand peak, which is precisely when interruptible service gets interrupted.

Operators in that position sometimes keep a modest diesel capability specifically for the controls and cooling load rather than for hashing, on the reasoning that protecting the asset is a different objective from protecting revenue. That is a reasonable position, and it comes with the storage regime described above attached to it.

Where the tank sits changes the whole job

Tank type is usually decided by whoever quoted the generator package, and it is worth a second look because the choice carries consequences for years.

Sub-base tanks sit under the generator as part of the package. They are simple, they occupy no extra footprint, and their capacity is limited by the physical envelope of the unit. For a site needing many hours of run-time, a sub-base tank is frequently too small, and the shortfall is discovered late.

Above-ground bulk tanks give you the capacity but introduce a separate structure, a transfer pump, a day tank in most configurations, and controls that keep the day tank fed. More capacity, more components, and more things that can be found in the wrong state during an outage.

Underground tanks solve the footprint and the temperature-swing problem and bring a regulatory and inspection burden that most mining operators would rather not take on. Rare on these sites for good reason.

Whichever you have, the component that most often fails quietly is the transfer arrangement between bulk and day tank. The generator runs on the day tank. If the transfer pump does not start, the engine runs for the day tank’s duration and then stops, with a full bulk tank sitting beside it. That specific failure is worth testing deliberately rather than assuming, because a routine generator exercise typically does not run long enough to exercise the transfer.

Test the thing you are actually relying on

A monthly no-load or light-load exercise proves the engine starts. It proves very little else, and it carries its own cost in wet stacking on a diesel unit run repeatedly at low load.

What the plan depends on is a chain: start, pick up load, hold it, draw from the day tank, refill the day tank from bulk, and keep doing that for hours. Testing the first link monthly and never testing the rest is how sites arrive at an outage with a surprise in it.

A load bank test exercises the engine properly under realistic load, and our guidance on load bank testing covers how that is run. The fuel-specific addition is to run it long enough that the transfer system has to operate, and to watch the day tank level while it does.

Refueling is a logistics problem before it is a fuel problem

Three questions settle whether a resupply will actually happen when you call.

Can the truck physically reach the fill point? Turning radius, overhead clearance, ground bearing on an unpaved approach, and whether the route is passable in the conditions that caused the outage in the first place. A site reachable in July is not automatically reachable in February.

Is there a standing arrangement? During a widespread event, suppliers serve existing contracted customers first. A site with no prior relationship is calling the same numbers as everyone else, at the same time, with no priority.

Who is on site to receive it? A delivery that arrives with nobody present to open the gate and witness the transfer is a delivery that leaves again.

Remote sites carry all three of these more heavily, and our guidance on remote site connectivity covers the related problem of knowing your fuel level before you are standing next to the tank.

Monitoring beats inspection

Manual level checks tell you where the tank was when somebody last looked. For a site running unattended, that is not sufficient.

Continuous level monitoring reported back with the rest of your telemetry gives you consumption rate as well as level, and the rate is the more useful figure. A tank draining faster than the load explains is a leak or a theft, and both are worth knowing about early.

Set an alarm threshold at the level where your supplier’s realistic response window still fits inside your remaining run-time. That threshold is a function of their lead time, not a round number, so work it out rather than picking twenty-five percent because it sounds prudent.

What to check on your own site this week

• Calculate real run-time at your genuine backup load, using the manufacturer’s consumption curve and usable capacity rather than nominal.
• Find out when the stored fuel was last tested, and if the answer is never, sample it.
• Drain the low point and look at what comes out.
• Confirm you have a named supplier with a standing arrangement, not just a phone number.
• Drive the delivery route and ask whether a loaded truck makes it in bad weather.
• Check whether tank level reaches your monitoring, and whether anyone would be alerted at night.

None of this is expensive. It is a half day of attention on the part of backup power that gets least of it, and it is the part that determines whether the generator you paid for does anything useful when the grid goes.

If you are specifying backup power for a containerized deployment and want the fuel side sized properly rather than assumed, talk to our team.