Pipe sizing is, of course, one of the main purposes and challenges of plumbing design. Everyone reading this article should already know that water supply fixture units (WSFU) are the basis for performing this task, coupled with friction and velocity limitations. Fixture units are converted to flow using Hunter’s Curve. Friction limitations are determined based on available pressure less the minimum pressure divided by system pipe length in hundreds of feet. This yields a friction factor in pounds per 100 feet of pipe. A very common friction factor might be 3 psi/100 feet.
Second to the friction factor is velocity limitation. In the Uniform Plumbing Code (UPC) the velocity limits are 8 FPS for cold water, and 5 FPS for hot. The IPC suggests 8 FPS for cold and 4 FPS for hot, although the verbiage is vague — at least in the 2009 edition that I possess. It may have been clarified in more recent editions. Some projects have acoustical criteria limiting velocities to something less than code, 6 FPS being very common, but I think this is overkill and will explain below.
The net result of these two criteria, friction factor and velocity, is that small pipe sizes are governed by friction, and larger pipe sizes are dictated by velocity as depicted in the flow chart that should be all but too familiar.
In theory, when piping is configured in a down-fed vertical supply, the friction factor for the vertical piping can basically be ignored due to static gain. Since pressure increases 4.33 psi every 10 vertical feet, or 43.3 psi/100 feet, all pipe sizes can be determined by velocity alone. However, I’m not sure how many plan examiners or inspectors would acknowledge this fact. You might be better served by sizing all pipes, both horizontal and vertical, with the same friction factor and velocity.
This is the widely held theory behind water pipe sizing. What I always find interesting is when the theory can be compared with reality.
On one of my hospital projects designed and built several years ago, we had reason to go back and install an ultrasonic flow meter on the domestic water supply. For this project the total number of fixture units was 3,685, which translates to approximately 500 gpm (in theory) and requires a 6-inch pipe (in theory).
This project was predominantly an outpatient facility with only 60 patient beds. When we measured the flow, we found the weekday peak to be 140 gpm and the weekend peak to be 70 gpm. This is a far cry from the calculated code dictated 500 gpm.
So why the big disparity? I hear people say on a regular basis that the code is a minimum standard, but I find that laughable. On every project I have visited after construction, I’ve watched the domestic pumps operate when the building is occupied, and I never see more than one of the triplex pumps operate at a time. So why is this?
I think there are two contributing factors at play here. First, I think Roy Hunter was more conservative than commonly thought when he was developing Hunter’s Curve. Second, water conserving fixtures are yet to be factored into the model codes.
For example, Hunter’s Curve suggests that there is a one-to-one relationship between fixture units and gpm for low flows of non-flush valve fixtures. I say “suggests” because the curve stops short of the origin at about 5 gpm and 5 WSFUs.
So the suggestion is that 1 WSFU is 1 gpm. Since a lavatory is 1 WSFU, and most codes limit the flow of public lavatories to 0.5 gpm, that is a 50 percent reduction that hasn’t been factored into the code. In residences with bath/showers, each of these carries 4 WSFUs, but since tubs are rarely used and showers are limited to 2.5 gpm or less, there is nearly a 40 percent reduction in flow from the allotted fixture units. These differences and others, of course, become cumulative.
Some time ago I was speaking with one of the hot water tempering valve manufacturers (I think it was Powers, but I’m not certain) about hot water flows it had measured in several different hotels and casinos. The largest hot water flow rate it measured was in the Bellagio, and that was 160 gpm. I don’t know the design flow for that system, but I can guarantee you it was three or four times that.
What do we do with all this information? We still need to satisfy the code with our designs, but we should keep this in mind when sizing piping and equipment. The code is not really a minimum standard; it is a conservative standard, so adding safety factors as I see so many engineers do is entirely unnecessary. The code needs to be updated to acknowledge water-conserving fixtures and the like.