Skip to content

The True Cost of Cheap Sand

Executive Summary: Every completions budget has a line for sand, and almost every one of them gets negotiated the same way. Price per ton, delivered. It is a clean number. It compares easily across bids. It gives procurement something defensible to put in front of finance.

It’s also the wrong number.


The True Cost of Cheap Sand

Sand is a commodity in the sense that matters least. Mesh distribution, crush strength, roundness, moisture spec: plenty of suppliers meet all of it, and a ton of in-spec 100 mesh from one mine performs about like a ton from another. Competing on grain quality alone ran out of room years ago.
What has not been commoditized is everything that happens between the mine and the blender. That is where completion performance is won and lost. It never shows up in a price-per-ton comparison.

What an hour actually costs

A frac spread sitting down costs somewhere between $10,000 and $20,000 an hour, all in. Permian sand sells for less than $20 a ton at the mine. It has been drifting down, not up.

Run those two numbers against a single well. Fifteen thousand tons of proppant is a sand bill of roughly $265,000. At $15,000 an hour, a frac spread sitting idle burns through that in about eighteen hours. The entire sand purchase for the well is worth less than one day of pumps standing still.

A supplier who shaves a dollar a ton off that bill has saved $15,000. That is one hour. Win the negotiation and lose a morning, and the whole exercise was a loss.

There is a second number hiding behind the first. Getting sand from the mine to the blender routinely costs more per ton than the sand itself. Procurement treats the commodity price as the price. It is less than half of what the operator actually pays.

What the upside is worth

Efficiency arguments get waved off as soft. They are not. The work content of a completion is fixed, so anything that raises pumping hours per day comes straight off the calendar, and everything else follows from that.

The table below runs a representative eight-well pad. Assumptions are stated underneath so any of them can be swapped for your own.

Improvement The arithmetic Value per 8-well pad
Two more pumping hours per day (16 to 18) Work content is fixed at 400 wall-clock pumping hours. At 16 hours a day that is 25.0 days. At 18 it is 22.2. 2.8 days off the pad. About $1.0 million in spread cost.
Four more pumping hours per day (16 to 20) Same 400 hours at 20 hours a day is 20.0 days. 5.0 days off the pad. About $1.8 million.
Cost per completed lateral foot $9.0 million of spread cost over 80,000 lateral feet falls to $8.0 million at 18 hours a day. $112.50 per foot drops to $100.00. A saving of $12.50 per foot.
Earlier first production Eight wells online 2.8 days sooner, at an early-life rate near 1,000 boe per day each, is about 22,000 boe pulled forward. Roughly $1.2 million of revenue arriving 2.8 days earlier.

Assumptions

Eight wells, 10,000-foot laterals, 50 stages each, 15,000 tons of proppant per well. Two hours of pumping per stage, simul-frac running two wells concurrently. Frac spread all-in cost of $15,000 per hour. Baseline of 16 pumping hours per day, against an industry range of roughly 12 to 22. Twenty-five tons per truckload, 40-mile haul, six miles per gallon, $3.50 diesel. Blended realization of $55 per boe.

The rows overlap and should not be added together. The first two are alternatives to each other, and the third and fourth both derive from the first. Each row is a separate way of pricing the same underlying gain in pumping hours.

Two hours a day is not an ambitious target. It is the difference between an operation that runs out of sand twice a shift and one that does not.

The real risk isn’t the price for a ton of sand. The real risk is the impact of 2-3 hours of downtime because your supplier missed their delivery or sand quality took down a pump.

The supply chain is the product

Once you accept that, supplier selection changes shape. The question stops being what a ton costs. It becomes how reliably a ton arrives, in spec, at rate.

Local supply with depth behind it. Distance adds failure points. A single-source supplier has no answer when a plant goes down or a frac calendar moves two weeks. Multiple mine sources close to the pad is what absorbs schedule volatility instead of passing it through to the operator.

Miles you never drive. Trucking miles are worth more than the diesel they burn. Every mile is driver hours consumed, road exposure accumulated, and dependence on a labor pool that has been tight for a decade. Cutting miles cuts cost, accident risk and exposure to driver shortage at the same time. The 42-mile Dune Express conveyor exists for this reason. It moves proppant across the longest leg of the journey without a truck, a driver or a road.

Minutes at the load-out. Load time at the mine is dead time, repeated thousands of times across a pad. Atlas turns a truck in roughly eight minutes. Much of the industry takes fifteen or more. That gap compounds into hundreds of truck-hours over a multi-well program, and none of it appears on an invoice.

Hours the clock does not bound. Atlas and Kodiak operate what Kodiak describes as the largest fleet of driverless big rigs deployed anywhere. As of the first quarter of 2026, 28 trucks ran 15 routes, carrying more than 450,000 tons across roughly 7,000 loads, logging over 23,500 driverless hours in that quarter alone. The fleet is scheduled to reach 100 trucks by mid-2027. Deliveries stop being bounded by one driver’s hours of service, and service continuity stops depending on who shows up for shift.

Optimizing the wellpad avoids NPT

Getting sand to location is just the start.

Opti Order monitors proppant inventory at the wellsite in real time and places orders automatically against live inventory and operational data.

Opti Dispatch schedules and routes those deliveries without human intervention. The point of both is straightforward. A sand-out should be a solved problem rather than a late phone call.

Everest is the wellsite delivery and handling system. It keeps more inventory on location, moves sand to the blender more cleanly, and rigs up faster without permitting delays. That pulls non-productive time out of both ends of the job.

PropFlow screens debris out of the stream before it reaches the blender. Rocks and foreign material damage pumps and fluid ends, and that damage arrives as an unplanned shutdown in the middle of a stage. Filtering at the wellsite protects the equipment and supports continuous pumping. It also removes the drying requirement, which widens the range of usable trucks to ordinary end dumps and belly dumps. That is one more pressure valve on the driver shortage.

The number that actually matters

Put the pieces together and an operator gets four things, all of which land in a completions engineer’s performance review.

Shorter total completion duration. Lower NPT. More pumping hours per day. Less equipment maintenance and fewer unplanned shutdowns.

None of those improve by shaving two dollars off a ton.

The better metric is cost per ton pumped, and better still, cost per completed lateral foot. Price per ton delivered measures what leaves the mine. Cost per ton pumped measures what actually made it into the well, on schedule, at the loading the design called for. Those two numbers diverge, and the gap between them is where cheap proppant stops being cheap.

Operators already apply this discipline elsewhere. Frac spreads and rigs get evaluated on utilization and cycle time, well beyond day rate. Sand has earned the same treatment.

The cheapest ton may cost you an expensive well.

Atlas Perspectives

Related Articles

Atlas Perspectives
Back to top