So what is the advantage of flowing more water in a flow test?
While the procedures for running a hydrant flow test are fairly well laid out, we have a step by step for us in the NFPA 291 standard, there are plenty of points within a test that create error.
Sources of error in a hydrant flow test include gauge accuracy, our ability to read the gauge accurately, and gauge position and placement among many other facets of measurement.
One way to help combat these sources of error, And to give us more confidence in the results, is to test with the flow rate that is beyond the flow demand of the system that we're working with.
As a starting point, NFPA 291 recommends that we flow enough water to achieve a residual pressure that is 25% lower than our tested static pressure, or to flow the total demand necessary for firefighting.
As an example, let's look at a theoretical test with the same equipment, but run at 2 different flow rates.
We'll go into conducting flow test and far more detail later on, but as a quick recap, here's how we run a typical two-hydrant flow test:
We first have our test hydrant, where we are monitoring the pressure at two different conditions. This is our static and residual hydrant, because we are measuring our static pressure and residual pressure at this hydrant. for this we’re going to refer to this hydrant as the test hydrant.
We then have our flow hydrant, which we are using to flow water and see how the water supply reacts when there is a flowing condition. on this flow hydrant we measure a pitot pressure so that we can estimate about how much water we are actually flowing during our test.
So the first step is that we open up and measure the static pressure from our test hydrant. Let's say in our test this static pressure is 60 psi. We establish this pressure as the static pressure add a no flow condition. let's chart that on our graph.
Our second step is that we open up the flow hydrant until we have a somewhat consistent flow coming through this flow hydrant. In this first example let's say yeah we only open things up a little bit. At this flow hydrant we measured the pitot pressure, and simultaneously take a pressure reading from our test hydrant. This pressure reading at the test hydrant is our residual pressure. In this example our residual pressure is 57 psi.
We then close out the test and very slowly close the hydrants and clean up our mess hopefully not hurting any of the landscaping nearby. Again we'll get into all of our tips and tricks and suggestions on flow tests a lot more detail later.
We can't just use this pitot pressure, however. We don't want to know the pitot pressure, we want to know the flow rate coming through our flow hydrant. So we have to convert it over to and estimated flow rate.
We take the pitot pressure combine it with a coefficient of discharge and convert that pressure over to a flow rate that we had coming from our flow hydrant. this is our residual flow rate. for this test let's say we have a coefficient of discharge of 0.9 flowing one 2-1/2 inch outlet with our 6 psi pitot pressure. That comes out to about 410 gpm. will cover this process in this calculation in more detail later.
Once we have our residual flow rate, which is 410 gpm, and our residual pressure from the test hydrant, which was 57 psi, we chart that on our graph.
Now, we use these two test points to establish our hydraulic relationship between them, by drawing a straight line on a log graph.
Now let’s say we run this same test, however, we fully open the hydrant using two side outlets instead of one.
We test our static pressure, still 60 psi, and chart that on the graph. Then we open our flow hydrant with two side outlets, and measure a pitot pressure of 9 psi with that hydrant now fully open. Our residual pressure with this much water flowing comes all the way down to 44 psi.
Converting the pitot pressure to flow, with two outlets, we get an estimated flow rate of about 1,010 gpm. We take this flow rate, with our residual pressure of 44 psi and plot that on our graph.
We can see here that this relationship is very similar. It’s the same water supply curve, so a test at any point along the curve should show a similar relationship, although theory doesn’t always perfectly align with the real world.
Now, let’s see what introducing a little error would have meant on these two tests. Just take, for example, the reading on a single gauge alone. if you've ran or witnessed a flow test you know that analog gauges like to bounce around. It's not uncommon to have a needle bounce between 2, 3 or 4 psi. Not only that, but if we have a 300 pound gauge that we’re trying to distinguish between a 57 psi or 58 psi reading, it’s extremely difficult to accurately read.
Let’s say that the gauge on the test hydrant is bouncing around, and we mistakenly read 58 psi residual when we should have read 57 psi. This is a 1 psi difference, or, in this case, less than a 2% error in our reading. Let’s draw out this new reading and see what impact is has on our available water supply calculation.
Here’s our new curve, with this slightly higher residual pressure. In-between the static and residual readings, or in the range that we would consider to be an interpolation, this error is still small. This error with a system demand at 200 gpm would have only made a 0.5 psi, or 1% difference.
However, while the error is minimal in interpolation, when we extrapolate points on our curve, our error is much larger. Let’s say our system demand is a Dry, Ordinary Hazard Group 1 system on a slope and our flow is about 800 gpm, now this error results in a difference of 3.4 psi, or nearly 7% error. If we get into a Dry and sloped Ordinary Hazard Group 2 system, or a storage system that creeps up around 1,100 gpm, now our error is over 6 psi off. That’s 13% off the real water supply.
The more and more we extrapolate beyond what our residual test point was, the more impactful any small error in our test becomes.
If we were looking at fire flow – our first reading at 57 psi residual would have us estimate the fire flow available to the fire department to be 1,670 gpm at 20 psi. Now, with our slightly higher residual pressure, we’re calculating fire flow available to a fire department to be 2,070 gpm at 20 psi. That’s a 400 gpm or 24% error! The further from our data we extrapolate, the more error we get from even slight deviations.
But let’s say we’ve got laser-focused eyes, our test equipment is all calibrated, our pitot gauge is perfectly centered, and we run a nearly flawless test. There’s still uncertainty. A calibrated gauge can still be off by an allowable margin. And our measurements are not perfect; they’re still an method to estimate a supply. The problem with flowing too little water is that any small deviation or uncertainty is compounded when we extrapolate data later on that’s far beyond our measurement range.
Now let’s take that same small error and look at it
Let’s say we again slightly misread the bouncing needly and overestimate our residual pressure again by 1 psi. So, at 1,010 gpm we read 45 psi instead of 44 psi. Let’s plot and draw this new curve. Now, this deviation has significantly less error than when we flowed less water. It’s because our tested range is much wider, and we are either extrapolating much closer to our tested range or better-yet we’re now interpolating.
That system demand of 800 gpm, which was off by 3.4 psi (or 7%) before, is now only off by 0.6 psi, or just over 1% error. That’s far less error than when we flowed a small amount of water.
Our fire flow estimates improve substantially too. Our original curve would estimate fire flow to be 1,650 gpm, but now, even with a mismeasured residual pressure, our fire flow estimate becomes 1,710 gpm, or being off by 3.6%.
Now do we always need to run a hydrant flow test down to 20 psi to find the fire flow? No, that’s not usually necessary, potentially harmful for the water supply, and in many cases not possible without a pump pulling more flow out of the system.
What NFPA 291 recommends is to flow enough water that our residual pressure is at least 25% below the static or what the flow rate is for firefighting purposes. This gets our test out far enough to give some confidence and overcome small deviations in the measured results.
Why is it important to flow more water during a fire hydrant flow test?
Flowing more water gives us greater confidence in the data that we pull from the water supply, and gives our results more tolerance from slight deviations or errors in the testing process.
I'm Joe Meyer, this is MeyerFire University.