Load is arrivals times dwell
160 cars leave the tunnel in the peak hour, 60 percent of them pull into a vacuum stall, so 96 arrive an hour. Each occupies a stall for 8 minutes, which is 96 times 8 over 60, or 12.8 stalls of load. That number is the same whether the lot has ten stalls or forty. Building more stalls does not reduce the load; it reduces the chance that somebody who wants a stall cannot have one.
The curve is flat and then it is not
At 16 stalls against 12.8 of load, an arriving car finds every stall taken 8.06 percent of the time. Add three stalls and that falls to 2.57 percent. Add three more and it is 0.56. Take three away and it is 18.45. The last few stalls before a target each buy a great deal and the ones after it buy very little, which is the mathematical shape behind the two complaints every vacuum lot generates: it is never full, except when it is completely full.
Reaching a 5 percent target at these numbers takes 18 stalls, two more than the defaults have.
Two minutes of dwell is worth three stalls
Dwell and stall count trade one for one in the load. Cutting the average visit from 8 minutes to 6 takes the load from 12.8 to 9.6, which at 16 stalls takes blocking from 8.06 percent down to 1.72. That is worth more than three extra stalls and costs no concrete. What moves dwell is mundane — mat racks and bins beside the stall so nobody walks, hose that reaches the far side without repositioning, and a layout that does not make backing out a three-point job.
What the model does not see
It assumes random arrivals and no queue: anybody who finds the lot full goes away. Real lots break both. Arrivals come in bursts because a tunnel produced them in a burst a few minutes earlier, which makes short-run blocking worse than the average says. And people do wait, sometimes in the exit lane, which is the failure the arithmetic is actually warning about rather than the one it models. Use the numbers for the gap between stall counts, not as a prediction of Saturday afternoon.
Questions people ask
How many vacuum stalls does a car wash need?
It follows from arrivals and dwell rather than from a rule of thumb. Cars an hour pulling in, multiplied by minutes each stays, divided by 60, gives the load in stalls. The stall count then sets the chance somebody finds none free. At 96 arrivals an hour and 8 minutes each the load is 12.8 stalls, and 18 stalls holds that to a 5 percent chance of none free.
How long does a car actually spend at a vacuum stall?
Longer than anyone guesses, which is why this page asks for a stopwatch figure rather than assuming one. Time it from pulling in to pulling out, including parking, getting the mats out and putting them back. The number matters more than almost anything else on the form, because dwell trades one for one against stall count in the load.
What does average stall utilisation tell me?
Less than the blocking figure does. Utilisation is the average share of stalls in use, and a lot can average 70 percent while turning people away regularly, because arrivals are lumpy. The number worth designing against is the chance an arriving car finds every stall taken, which is what the table on the page walks across stall counts.
Is it cheaper to add stalls or reduce dwell time?
Usually to reduce dwell, and by a wide margin. Two minutes off an eight minute visit is worth about three stalls at typical arrival rates, and it comes from mat racks, hose reach and a layout that does not make backing out difficult rather than from concrete. The page prices both so the comparison is visible rather than assumed.
How much ground does a vacuum lot take?
The page multiplies stalls by a rectangle you supply for stall width and depth plus aisle, and the answer is a first cut only. It carries no allowance for the drive from the tunnel exit, no turning radius at the row ends, no landscaping, setback or drainage. What may actually be built, how it drains and how much of a site can be paved are questions for a civil engineer and your local planning authority.