Recommended FLOW for rotary espresso pump

Need help with equipment usage or want to share your latest discovery?

#1: Post by DolceVitaBaby »

Hi, I have been doing some research and it seems I own some Procon Pumps that may be TOO powerful. They pump 100 gallons per hour! I'm seeing references to 200 liters/hr max. Who knew. What is the recommended liters/hr for an espresso pump? What is the minimum? Is this posted anywhere? Don't think so!


#2: Post by djmonkeyhater »

Can you give us a little more context on what you are trying to do? Are you modding a machine? Have a pump in your basement/garage that you'd like to use for something?

The volume of water consumption for an espresso machine is pretty modest even if it being used in a high-volume commercial application so almost any rotary pump could output something on the level of 10-50x what you would need. I have one in a box on my workbench and it is rated at 100 l/hr or 26 gallons. It would be standard for almost all sizes of machine that take rotary pumps whether used twice daily at home on a small single or nearly constantly on a busy 4 group machine.


Capuchin Monk

#3: Post by Capuchin Monk »

Rotary vane pump for espresso machine I saw on ebay shows as 174 l/hr @ 7 Bar.
Here's a link on Fluid-o-Tech pump spec.

User avatar

#4: Post by shadowfax »

200 L/h is pretty high for an espresso machine; I have a 1 group Elektra that came with a 150 L/h pump, and that was just sick: the darn thing practically cavitated during boiler fills, as it would suck so hard on my low-flow-rate carbon filter. I ended up having the liners and pins in the pump replaced to reduce it to a 70 L/h flow rate.

Of course, anyone who does the math will realize the minimum flow rate you want at 9 bars: ~90 mL (3 oz.) in 30 seconds, or about 3 mL/s. Since there are 3600 seconds in an hour, you need a pump that will produce 3 mL/s * 3600 s/h * 1/1000 L/mL = 10.8 L/h. So, pretty much anything out there is ridiculously overpowered, but for good reason: in a commercial environment, you have to be able to feed multiple groups AND the boiler autofill without taking significant pressure drops. In order to do that, you have to have a very high base flow rate.

So, rotary pumps have a lot of potential power, and they are regulated by spring-loaded bypass valves that open and circulate output water back to the inlet (basically spinning water in a circle in the pump) when the desired pressure is reached. So any espresso machine with a rotary pump has its bypass valve open almost 100% of the time it's on, and the actual flow rate you see is determined by the flow rate that the group jet allows at the given pressure you have set on the pump. That is only a fraction of the flow the pump delivers, but it's what you see when you brew.

So technically, a 100 gph (378 L/h) pump *might* work OK for an espresso machine, assuming it has a PROPER bypass valve on it that can relieve adequate excess flow. It is a bit like using a hammer to put a thumbtack into a cork board, but it might work. Don't blame me if you shear a fitting, though.

Back to my bit about liners and pins: you might be able to contact a procon distributor and locate someone that works on Procon pumps and get them to change out the pins and liners to make its flow-rate much smaller (I would suggest 100-150 L/h for a commercial one group, less for a home rotary machine). Fluid-o-Tech USA charges about $10 for this conversion, IIRC, or $30 if you want a full refurbishing of the pump. I would anticipate something in that neighborhood from procon, if they make liners small enough for your pump.

Anyway, that's a general explanation of the factors that go into rotary pump flow rates for espresso machines. If you want something in real detail or some direct help, as Wesley said we'd need more specific information. Hope that helps.
Nicholas Lundgaard


#5: Post by Billc »

Pumps are rated for flow without things attached. Normally you sacrifice volume for pressure. So that pump you have is rated for XX gal/hr at some pressure (and temperature). The higher pressure the lower the actual volume the pump can output. Additionally if there is any restriction i.e. pushing water through small tubes, there is also some pressure loss. The pressure loss is overcome by a reduction in volume. Very interesting to note that this is the same thing that happens at the portafilter. Since flow is directly dependent upon pressure, lightly tamped coffee allows more flow and the pressure is decreased, tightly tamped coffee restricts water more resulting in higher pressure.

You really have to look at the manufacturers specs. to really determine if the pump is over rated for the machine. In some cases a pump that is rated for 700 L/hr may only achieve 10 L/hr at 9 bar.

Bill C

User avatar

#6: Post by shadowfax »

Billc wrote:You really have to look at the manufacturers specs. to really determine if the pump is over rated for the machine. In some cases a pump that is rated for 700 L/hr may only achieve 10 L/hr at 9 bar.
Very true and good point. Trying to bring my original comment to bear on your advice, I think this document from Fluid-O-Tech paints a pretty accurate picture of your average espresso machine's rotary pump's ability to produce flow at pressure. This is for a compact F-O-T pump like the one found in the GS3. I believe the GS3 is an MA074, that is a 70 L/h brass pump with a balanced bypass valve mounted. This pump is a lower-flow pump than you'd see on your average machine, particularly multigroup ones, from what I've seen, but typical of smaller 1-group 'home' machines that sport rotary pumps, such as the QM Vetrano, La Spaziale S1, Izzo Alex, etc.

As the graph shows, the GS3 pump in the US (60 Hz AC/1725 rpm, C' on the graph) can output ~70 L/h @ 16 bars.

A 150 L/h pump (again, US, E' on the graph) can put out ~170 L/h at 16 bars. As I said before, from what I understand of rotary pumps as they are used in the espresso machines, their potential flow is all but entirely untapped 99-100% of the time, and instead the balanced bypass on the pump (for the uninitiated, this is the screw with the lock-nut that points out at a 90° angle to the inlet line of the pump) regulates the pump so that it holds local pressure very roughly constant, at least at low output flow rates. In my experience, seems like you see a pump put out (upstream of the gicleur, as measured by a machine's pressure gauge) about 0.5 bars less pumping against 'nothing' (no portafilter in the group) vs. pumping with no flow (i.e. against a blind filter).

Of course, at the top of the puck is a whole other story, as you say, because things get more complicated when you put flow restrictors in a series and measure in between them and throw in the added curve ball that there's a transition from pumping a compressible medium (the air in the tube between the 3-way valve and the puck) to pumping a relatively incompressible medium (the water that pushes the air out). But it seems to me that appreciating and tweaking what's going on there is mostly a matter of changing your pump's bypass setting or changing the gicleur and/or your tamping/dosing/grinding. Changing your rotary pump's flow rating among the above doesn't seem likely to change much of this, as, again, every one of the pumps on the graph above is "spinning its wheels" to some degree or another when it's got its bypass at 9 bars and you're pulling a shot of espresso.

Getting back to the topic at hand, I'm pretty sure that if you get a Fluid-O-Tech or Procon pump rated for 30+ L/h, you will be able to brew espresso at one group at a time with no trouble at all. However, with those lower-flow pumps (I am thinking the 30-70, and maybe the 100), if the boiler autofill kicks in on you, or another barista flushes at another group during your shot, you will see a potentially large pressure drop on your shot. So the big multi-group machines put high-rated 150, 175 L/h pumps on their machines. So, your minimum is going to depend on the number of groups that your machine has, and to some extent whether you expect the autofill on the steam boiler to ever kick in during the shot (some nice espresso machines have controls that prevent this).

What I really have a poor handle on is "how much is too much?" I would think it'd be the point where the flow rate of the pump was so high that you couldn't set the bypass to hold as low as 9 bars during operation where the net output is ~3 mL/sec. I don't know at what point a rotary pump does that, but I'd guess a 100 gph pump might work, *if* it has a balanced bypass valve on it. If you don't have a bypass, don't even bother. If you have a 'regular' bypass, I think those work but are a lot more sensitive to inline pressure fluctuations, which might make a huge difference with a pump like that. The other thing that I can see being a real problem is cavitation. If you have a filter or some narrow tubing upstream that is rated for less than what the pump wants to put out (not necessarily its 'rating' so much as the output rates that it sustains at the pressure you set it to and the rate it can push through the fattest opening it's pushed out of, usually the steam boiler fill line), you risk cavitation. Seems like a bad idea that can damage things.

Does that sound right, Bill? That's what I've gleaned from my own tinkering, looking at these graphs, and talking to EricS and rotary pump vendors in conjunction with the pressure profiling project. Apologies to the OP if this is way more technical than you were looking for...
Nicholas Lundgaard

DolceVitaBaby (original poster)

#7: Post by DolceVitaBaby (original poster) »

Whew, I got the full monty on espresso pumps guys, and thanks.

User avatar

#8: Post by HB »

Follow-on discussion split to Espresso pour speed of heavy vs. light tamps.
Dan Kehn

User avatar

#9: Post by dsc »

Hi guys,

isn't flow rate only affecting the ramp up speed? I mean if you set your pump/OPV to produce 9bar you will get more water circulating in the bypass on a high rate flow pump than on a low rated flow pump, but I doubt there's much difference on the water-through-puck level. High flow/fast ramp up will definitely slow the pour down due to fines migration which will happen faster, so one will have to grind coarser to get the same timings (as on the slower ramp up pump).

Feel free to shoot me down if you think the above is wrong.



#10: Post by darrensandford »

I would expect the flow at the puck to be controlled by the gicleur (or some other restriction in the machine), rather than the flow rate of the pump?