GPM and PSI Selection by Task Reference
Why this matters
A tech who only thinks in PSI picks the wrong rig for a job as often as the right one, because pressure alone does not tell you how fast a job gets done or whether the pump can even sustain that pressure on the water supply available. Too little GPM against a surface cleaner rated higher and you get streaking no amount of extra passes fixes, because the tool needs a certain flow to rinse ahead of itself as it spins. Too much GPM demand against a garden hose spigot that cannot keep up and the pump cavitates, running rough and wearing itself out on a job that looked like a simple hookup. Both PSI and GPM have to be picked together, against the task and against the water supply actually on site, or the rig that looked adequate on paper is not adequate on the driveway.
PSI and GPM measure two different things
PSI (pounds per square inch) is the force behind the water at the nozzle. GPM (gallons per minute) is the volume of water moving through the system. Force does the cutting and lifting; volume does the rinsing, the coverage speed, and the flushing of loosened soil off the surface before it resettles. A trade heuristic for comparing two rigs' overall cleaning capacity is cleaning units, PSI multiplied by GPM: a 4 GPM machine at 4,000 psi runs about 16,000 cleaning units, while an 8 GPM machine at 3,000 psi runs about 24,000, meaning the lower-pressure, higher-flow rig moves more total cleaning capacity even though its PSI number looks smaller on the spec sheet. Treat cleaning units as a rough comparison figure for shopping between rigs, not a precise physics measurement, since it says nothing about tip angle or standoff distance, both covered in the [Nozzle Tip Selection by Surface and Task] article.
Task-based selection
| Task | PSI | GPM | Why |
|---|---|---|---|
| Softwash rinse (house, roof after chemical dwell) | Near-zero to about 500 | 4 or more | Flow rinses the chemical off; pressure is not doing the cleaning here at all |
| Wood decking and fencing, light pressure rinse | 500 to 1,200 | 4 | Force stays low enough to protect grain; see the nozzle tip guidance for angle |
| General concrete, wand only, no surface cleaner | 3,000 to 3,500 | 4 | Standard residential belt-drive rig covers this comfortably |
| Concrete with a surface cleaner, moderate area | 3,000 to 3,500 | 4 to 5.5 | GPM must clear the surface cleaner's rated flow range to avoid streaking |
| Concrete with a surface cleaner, larger driveway or small commercial lot | 3,500 to 4,000 | 5.5 to 8 | Higher GPM covers more square footage per hour without slowing the pass to compensate |
| Vehicles, outdoor furniture, delicate material | Under 1,200 to 1,500 | 4 | Wide tip and distance still do most of the protecting; see the nozzle reference |
| Brick, block, stucco | 1,200 to 1,800 | 4 | Wide enough tip at this pressure avoids mortar joint erosion |
These figures assume the 4 to 8 GPM belt-drive pump class this trade runs day to day, not an industrial rig. A machine's actual rated output, stamped on its data plate, governs over any of these figures.
GPM has to match the attachment, not just the pump
A surface cleaner is built around a rated GPM range printed on its own housing or spec sheet, because its cleaning action depends on the spinning bar rinsing each strip of concrete before the bar comes back around. Run a surface cleaner rated for 5.5 to 8 GPM off a 4 GPM pump and the bar outruns its own rinse water, leaving faint streaking in a pattern that follows the bar's rotation, no matter how many passes you run over it. This is an attachment-to-flow mismatch, not a technique problem, and it does not get fixed by working the surface cleaner slower; it gets fixed by matching the pump's GPM to what the tool is rated for, or dropping to a wand and a hand-held tip instead.
Pump cavitation: the water supply has to keep up with the pump
A pump can only deliver the flow its water supply provides. A standard garden hose off a municipal spigot commonly delivers somewhere in a range of roughly 9 to 17 GPM under ideal conditions, but real delivered flow is throttled by hose length, hose diameter, and how much pressure the spigot itself is actually running that day, so the number has to be checked on site rather than assumed from a hose's rated maximum. Ask a pump to pull more GPM than the inlet water can supply and it cavitates: pockets of vapor form on the suction side, the pump runs rough with an audible knock, output pressure surges and drops instead of holding steady, and sustained cavitation wears pump valves and seats faster than normal duty does.
Field test: time how long a 5-gallon bucket takes to fill straight off the spigot, with nothing else drawing from the same line.
- Bucket fills in 50 seconds: 5 gallons divided by (50 seconds divided by 60), which equals 6.0 gallons per minute of actual supply.
- Compare that to the pump's rated GPM. An 8 GPM pump against a 6.0 GPM supply is short by 2.0 GPM, 25 percent under what the pump wants to pull.
A supply that comes up short does not mean the job is impossible; it means the machine has to run within what the supply can actually deliver; either drop to a lower-GPM tip and pace, or bring a larger-diameter supply hose and confirm the improved fill time before the pump ever engages.
Splitting flow between two operators
An 8 GPM skid or trailer unit can run two wands off a single manifold, but the two guns share the pump's total output, they do not each get the full rated flow. Two operators running simultaneously off an 8 GPM pump split roughly 4 GPM each when both triggers are open, which drops each gun back into the single-wand residential band from the task table above rather than the higher band the pump's total rating suggests. That is fine for two techs each running a wand on siding or a fence at the same time; it is not enough for either of them to run a surface cleaner rated 5.5 GPM or higher on their own gun while the other trigger is also open, because neither one is actually pulling the pump's full rated flow in that configuration.
What changes the answer: private wells and long hose runs
A municipal spigot is not the only water source you will show up to. A private well system's pump has its own rated output, often well under what a municipal line delivers, and a homeowner rarely knows that number off the top of their head. Ask, but run the bucket test regardless of the answer, since a well pump can be undersized for its own house even by the homeowner's account. Where the test shows a shortfall, the fix is the same as on a municipal line: drop to a lower-GPM plan or bring a larger-diameter supply hose, not push the pump past what the source provides and accept the cavitation risk.
Hose length and diameter change what arrives at the gun even when the source itself is adequate. Friction inside the hose costs pressure over distance, more on a narrow-diameter hose than a wide one, and more over a long run than a short one, which is why a rig spec'd at 3,500 psi at the pump does not necessarily deliver 3,500 psi at the tip on a 150-foot high-pressure run. For any run long enough to matter, check delivered pressure at the gun itself with a gauge rather than trusting the pump's rated output as the number the surface is actually receiving, particularly on a job where the task table above puts you close to a surface's tolerance limit rather than well under it.
Worked example: picking a rig and confirming supply for a two-car driveway with a surface cleaner
The job is a moderately soiled two-car concrete driveway, planned with a 20-inch surface cleaner rated 4 to 8 GPM. The shop's mid rig runs 5.5 GPM at 3,500 psi, comfortably inside both the surface cleaner's rated range and the task table's concrete-with-surface-cleaner band.
Before connecting, the bucket test ran on the customer's outdoor spigot: 5 gallons filled in 34 seconds, which is 5 divided by (34 divided by 60), equal to 8.8 gallons per minute. That comfortably clears the pump's 5.5 GPM demand, so the pump's own inlet screen and the supply hose diameter were the only remaining checks (a kinked or undersized supply hose can throttle flow below what a wide-open spigot alone would suggest, so the hose actually connected got the same rough check by feel for kinks and a look at its diameter, not just the spigot itself).
The surface cleaner ran its full pattern with no streaking, confirming the flow match held in practice, not just on paper.
Verify before the job starts
- Confirm the task's target PSI and GPM range from the table above, or from a specific product's rated pressure and flow.
- Confirm any attachment (surface cleaner, specialty nozzle) is rated for the pump's actual GPM, not just compatible with its pressure.
- Run the bucket test on the actual water source for the day and compare against the pump's rated demand.
- Where supply falls short, adjust the plan (lower-GPM tip, larger supply hose, or a different spigot) before the pump runs, not after it starts knocking.
References
- See related: Nozzle Tip Selection by Surface and Task
- See related: Soft Wash Skid Setup and Tank Mixing
- Manufacturer data plate and owner's manual for the specific pump's rated PSI and GPM
- Surface cleaner manufacturer specification sheet for its rated flow range