Water Pressure Boosting System

Boosting the pressure of water can be useful for many reasons. In my case, I use water to cool a pair of high current electromagnetic coils. The cooling pipes have a fixed size that cannot be increased. By boosting the water pressure, I can increase the flow of water through the pipes and thus I directly increase the power capacity of my system. Higher pressure can also be useful for better cooling any electrical or heat generating system, for getting water to go a long way uphill, or for stronger spray jets in irrigation or showers. If you are interested in boosting your home water pressure, there are a lot of plumbing considerations that will be relevant for you which I don't know about and don't go into here.

In this instructable, I use a 3/4 horse power pump from Flotec, but I will also discuss pump selection more generally. It is distributed by Northern Tool, among others. This is a specific kind of water pump designed to give a large pressure increase by stacking multiple compression stages one after the other. Multi-stage pressure boosting water pumps typically deliver 100-200 psi above inlet pressure, and several can be placed in series, provided the total pressure doesn't exceed their rated "maximum total pressure", in this case 315 psi. By comparison, other centrifugal pumps deliver great flow rates, but don't usually increase the pressure by these amounts. For example, this pump from Star Water Systems provides 4500 GPH (eight times the Flotec) but only 48 psi. For really high pressures, linear piston based designs are more common. This piston pump provides 1000 psi for example at 180 GPH, and this one can hit the 10,000 psi mark.

If you're looking for pressure boosts of a factor of 2-4 for flow rates below 1000 GPH, you may be able to use a pump very similar to the one I used, but even if you're not sure what you need yet, this instructable should help you decide that as well.

Figure out what your flow rate will be. Do this by setting up your apparatus sans booster pump, and emptying into a 5 gallon bucket instead of your return or drain. Turn on at a known pressure. (There should be a gauge somewhere to tell you your building pressure. Mine is 52psi. If not, connect a gauge near where the pump will attach). Now measure the time it takes to fill in seconds. Five divided by this is your flow rate in gallons per second, and multiply by 60 to get gallons per minute or by 3600 to get gallons per hour.

Now, if your apparatus doesn't suffer from turbulence then the flow rate after boosting will scale with pressure. The Reynolds number can tell you this, provided you've also avoided right angle turns, choke points, or other flow impediments. Re = 10^6 * dh / v, where v is the velocity in m/s that you can get based on the flow rate you calculated, dh is the hydraulic diameter (normal diameter for a round pipe, side length for a square) in meters, and the factor of a million comes from the kinematic viscosity of water at 20 degC. Note that this actually changes significantly with temperature however. If Re is below 2300, you are in the laminar regime, where flow is non-turbulent, and flow rate scales linearly with pressure. If your flow isn't laminar, pressure boosting is probably not the right strategy for your application. On the other hand, if the booster isn't pushing you close to turbulent, then you can probably make further gains with a bigger boost.

In my case, I get the following:

1. volume flow = 0.125 GPM at 60psi,
2. dh = 3mm
3. A = 9mm^2 (pipe area)
4. v = 0.125 GPM * 1min/60sec * 4L/gal * 1m^3/1000L / 9mm^2 *1,000,000mm^2/m^2 = 1.9 m/s.
5. Re = 1,000,000 * 0.003 / 1.9 = 790 < 2300, laminar.
6. With the booster, pressure goes to 210psi, so Re goes to 2700, the transition region, possibly turbulent, but close enough that its probably okay.
7. I'm heating the water a lot (70C), so Re will go up by another factor of 2.5 and I will definitely have turbulence near the outlet of my coils. So I probably won't get a pure scaling of flow rate, and a larger pressure boost probably won't help more.

A note about units- I prefer to work with psi, pounds per square inch. An atmosphere (100 kPa, 760 Torr, 1 bar, 30 inHg) is about 13 psi. For water pumps, some prefer "feet of head" which literally tells you how high uphill you can run a pipe before the water pressure drops to zero and the flow ceases. For Mercury, which is a metal liquid 13.5 times as dense as water, one atmosphere only pushes 30 inches or 760 mm uphill. For water, you guessed it, an atmosphere pushes 13.5 times as far, or 34 ft. This means that if you had a 35 ft straw and you tried to put your finger on the top and lift it out of the ocean, the top foot closest to your finger would be under vacuum, and the water would only rise the first 34 ft. Unless you live at altitude like me. In boulder it would rise 31 ft. Anyway the upshot is that pumps also come rated for "feet of head", of which there are 2.3 for every psi. Here's a handy converter.

A note about power: pressure times volume flow gives units of power, since pressure times volume gives energy and energy per second is power. The units require a bunch of conversion, but it is possible to multiply the pressure gain and volume flow rate of a pump to get the power. If I do this for my 150 psi, 780 GPH pump, I find more horses than advertised:

150psi * 10^5 Pascal / 13psi * 780 gph * 4 L/gal * 1 hr / 3600 s * 1m^3 / 1000L * 1 hp / 750 W = 1.33 hp.

This is because no pump can actually deliver maximum pressure boost at maximum flow rate. If we really want 780 gph, with 3/4 hp we can only get .75/1.33 * 150psi = 85 psi boost, but this neglects losses. Looking back at the specs I find the actual pressure boost at max flow: "13GPM @ 70psi" (13*60=780GPH) This is important to keep in mind for pump selection, which we'll get into more next.

The 3/4 hp model I chose from flotec allows me to boost by 150 psi and from 15 GPH to about 50 GPH. 1/2 hp and 1 hp models are also available, with pressure boosts of 110 psi and 190 psi respectively. There's plenty of other reasonable looking manufacturers, grainger has a nice list here. One thing to keep in mind is that these pumps can't be run at zero flow, because they rely on the water for cooling. I couldn't find a minimum flow rating for the flotec pump, but page 5 (bottom right) of its manual explains that in a spray nozzle application, nozzles must be "weeping" so that they provide flow when not spraying. I infer from this that the required minimum flow rate is rather small compared to the maximum flow, probably only a few percent. If your application requires much less than 50 GPH, you could consider adding a bypass pipe that gives the water another path from high pressure back to your drain to increase the flow rate through your pump. A long roll of 1/4" PETE tube with 350psi rating should do the trick, its length can be tuned for any desired flow rate. As long as the flow allowed by the bypass stays below 10% or so of the pump's max flow, the pressure will not be substantially reduced. This means the flow along your desired path will not be reduced by the bypass.

The pump must be firmly mounted. Failure to do so will cause it to shake and wear out its bearings during operation. Preferably the mount connects the pump to concrete which can absorb the vibration, such as the foundation of your building. If there is already a unistrut frame anchored to the cement floor, as in my service corridor, this is an ideal place to connect the pump. If you must settle for screwing the pump to a piece of plywood, make it larger than need be and screw weightlifting plates to it.

If you're using a flowtec, it comes with 115 and 230 volt options. If you have access to 230, it's a better choice because the current will be lower. At 115 you may be close to 15 amps, (even with low water flow and thus low power draw, you'll get close to peak current at least when you first plug it in), potentially throwing your breaker or overheating wiring. If you have access to 120 volt three phase, you can make 230 volt 2 phase out of it just by connecting only two of the three phases. It will really be 208 volts, but that is normally close enough for most equipment. If you do this, make sure you cap off the unused phase and have no copper exposed when you turn it on.

To be more specific, lets say there is an outlet available in your building that looks beefier than normal and has a label saying 230 or 3phase. Go to mcmaster and find the corresponding plug. In my case it looks like this. Count the number of tabs. If there are three, than one is a ground and you have two phase power. If there are four, then one is a ground and you have three phase power. Now find cable with a low enough gauge for your current draw. You can strip the wire like normal and connect it to the plug with a screw driver. On the other end, strip and connect only two of the phases and ground as directed in your pump manual (page 6), and put a cap on the third phase. Now if your pump needs three phase but you have two, that's a different, more challenging story.

If you are using a larger model at 1 hp or above, you may need to have an electrician wire an outlet back to a breaker with a larger current limit. Typcial breakers stop at 20 or 30 amps.

Now for the actual business. Depending on which horsepower pump you choose, the inlet and outlet will have some NPT size that you will need to match. For my application, I chose to adapt to hose right away. This yields a less permanent installation, but with high enough pressure rated hose, it's no sacrifice of quality and can even be better than pipe for avoiding right angles or dissipating water hammer. Use an appropriately sized hose barb to mate to the pump. It should be male NPT threaded on the pump side with a size matching the pump. The 3/4 hp Flotec uses 3/4 size NPT. The other side is sized to the inner diameter of your tube.

Keep pressure ratings in mind for everything. The brass fittings I used everywhere are good for 250 psi, but if you're going a bit higher, stainless steel will get you to 300 psi. Beyond this I don't think you'll find plastic tubing that can work anyway, so you'll have to do some more serious plumbing with actual metal pipes.

A note on NPT threading. You'll need teflon pipe tape, and you need to wrap it in the right direction. If you didn't know about direction, that might explain some of the times you've tried to use teflon tape and it has still leaked. This how-to gets it right.

The electronic system that I cool with this system is a 300 Amp, 25 turn helmholtz coil suitable for generating 350 Gauss at the center of a 6" cube vacuum chamber. The coil is wrapped with epoxy coated square copper pipe, since the square profile allows tight packing. However, this means that I need to have a custom solder joint somewhere. If you have this as well, make sure to use 95-5 tin-antimony solder as it sustains higher pressure. Close to boiling water temperature (depending on your load, you might want be heating your water close to boiling like me) this solder is still rated to 300psi.