Off-Grid Solar Water Pumping Using Multiple Espresso Machine Pumps (ULKA Type)

Hello everyone,

I'm working on a modular and low-tech solution to pump water off-grid using solar panels and components that are easy to find and maintain.

The idea:
Use 3 or more ULKA-type espresso machine pumps (the kind used in coffee machines) in parallel, driven in sequence using a shift register or PWM controller. Each pump would be powered directly by a 24V or 36V solar panel, without batteries or inverters.

Why this matters:

I need to pump water from a hydraulic ram pump's waste outlet uphill (~40m head).

Flow rate needed is low (around 4–8 L/min), just enough for grazing livestock.

Redundancy: if one or two pumps fail, the others keep working.

Using small piston pumps avoids expensive high-pressure brushless pumps.

Cheap solar timer or vibration sensor could sync the pulses with the ram pump cycle.

I've tested ram pumps and even considered a Pelton wheel turning a gear pump, but this ULKA idea could be simpler and scalable.

Questions:

Has anyone tried something similar?

Are there better compact piston pumps for water (durable, 15+ bar)?

Do you think this is more efficient than a second ram pump?

Happy to hear any feedback or improvements. I’m known online as “udos46” and love mixing old analog tech with modern low-power ideas.

Thanks!

I would never have thought that espresso pumps ran off of anything but AC, 220v Euro or 120V U.S.
Marine/RV pumps come in 24V style, more expensive than your espresso machine, but much higher flow, therefore more amps, my bet. Usually diaphragm types.
The 12v ones are maxxed out at your 40m head of water.
I use them around the property to supply running water from 1000L storage totes. Small solar panel, deep cycle battery. Works very well.
Your idea of 36V solar panel with voltage controller is a cool idea.

bruno motta wrote:Hello everyone,

I'm working on a modular and low-tech solution to pump water off-grid using solar panels and components that are easy to find and maintain.

The idea:
Use 3 or more ULKA-type espresso machine pumps (the kind used in coffee machines) in parallel, driven in sequence using a shift register or PWM controller. Each pump would be powered directly by a 24V or 36V solar panel, without batteries or inverters.

Why this matters:

I need to pump water from a hydraulic ram pump's waste outlet uphill (~40m head).

Flow rate needed is low (around 4–8 L/min), just enough for grazing livestock.

Redundancy: if one or two pumps fail, the others keep working.

Using small piston pumps avoids expensive high-pressure brushless pumps.

Cheap solar timer or vibration sensor could sync the pulses with the ram pump cycle.

I've tested ram pumps and even considered a Pelton wheel turning a gear pump, but this ULKA idea could be simpler and scalable.

Questions:

Has anyone tried something similar?

Are there better compact piston pumps for water (durable, 15+ bar)?

Do you think this is more efficient than a second ram pump?

Happy to hear any feedback or improvements. I’m known online as “udos46” and love mixing old analog tech with modern low-power ideas.

Thanks!

I would love more detail for context. usually a hydraulic ram would be used to pressurize a lower volume of water then it takes in and increase its pressure to provide higher pressure water for say irrigation or home use. Why do you need to pump the outflow uphill? Why not pump it once using a better pump that is solar driven? Lots of high pressure solar pumps available for less than the cost of 3 espresso pumps and designed for the job? I don't want to poke holes just trying to understand.
Cheers,  David

Interesting idea and very common. I had this problem before and came to a solution.
Problem:
Water below the house of potable (drinking water) quality running at 16L/min. Animals need water in the fields 200m away on either side and 10m above the spring. The spring is 10m above the house and 100m away from the house.
The setup:
All water from the spring is collected in a sanitised installation and send to 10000L tank by the house. The water overflows from the tank (most time at full flow) to a drive tank for a ram pump. This is made from a plastic street gully with a rain proof lid. I only used this as it is the cheapest solution as I needed at least a foot of space in diameter. The drive pipe to the ram pump is fitted to the bottom of the tank and reaches into the tank with a 1 inch thread. The drive pipe to the ram pump is 12m long and 1" bore stainless. This will not flex and diminish the ram effect. The ram pump is made entirely of stainless steel, but the expansion vessel. The supply pipe from the pump feeds a tank on top of the farm that supplies all fields above the house with animal water, everything below the house is supplied from the storage tank by the house direct.
As the animal water for the field below the house is drawn from the tank that feeds the ram pump, the ram pump water supply will be interrupted. That is where the extended thread into drive tank (as mentioned above) comes into action. On the end thread inside the tank is a elbow facing upwards that has 1.5" modified checkvalve on it that is fitted in its blocked direction. The modification is such, that an eyelet has been fitted to the plunger, allowing it be opened. On the eyelet is a piece of wire that connects to the handle of petrol can as a float.
Here is how it woks:
When there is enough water filling the drive tank, the petrol can, working as a float, opens the reverse checkvalve and lets water into the drive pipe, operating the ram pump, as it will have no resistance if the pipe is empty. If the checkvalve in the pump is good and has retained pressure and the supply pipe is full and not draining the back pressure, all will work.
Now, when the drive tank lacks water, the float will lower and stop the flow into the pipe. The valve will be sucked onto the seat, and if there is seepage the remaining water in the drive tank will keep the drive pipe full as long as there are no air leaks into the pipe on the way to the pump.  
Using solar power for pumping is less complicated. All you need is a 12V 60W caravan shower pump and 2 100W PV panels, one facing  SSE the other SSW and you will have 60W some of the day, feed the output of that into the delivery pipe via a check valve for protection.

Hope this might help. All the best,
Cesar.  

Further to my last posting above with regards to self starting ram pump.

I use a lot of boat bilge pumps for hydroponics and moving water in my greenhouse. At less than 15.00 dollars they are affordable. The head is only 3 meters but would a series of 3 get you there? Some have a float valve, so installed in a small reservoirs you could have them pump the water when it became available.

I believe these ULKA type pumps do require alternating current.  I'd suspect an inverter would suffice, though.  There's not much to these, certainly no sensitive electronics which would be finicky about the exact wave form of the AC current, if I understand them correctly.  They may not even be all that picky about the exact frequency (see below).

In the summer of 1994 I had an internship in Essen, Germany, working for a filial of Johnson Controls.  This facility mostly made gas and water valves in an old panzer factory that had somehow made it through the war unscathed, and also made the electronic control modules for the gas valves (as I recall, the water valves were all manually operated).  It had the old-fashioned sawtooth roof for natural daylighting, and the floors were reputed to be over a meter thick of reinforced concrete, but the heaviest thing moved around the floor while I was there was probably the coin op beer machine, which they were removing from the factory floor that summer (the end of an era!).  This wasn't long after the Berlin Wall came down, and there were a lot of changes happening, and maybe not all for the best.  The number of obviously intoxicated people I saw that summer I could count on one hand (most of them panhandling at the main train station), and the number of truly obese people I encountered was exactly one - and I spent a lot of time on public transport going hither, thither and yon throughout this city of some 700,000 souls.  I haven't been back since, but I wonder what it's like, now?

That summer, I spent a fair bit of time working in the quality lab, running valves on a flow bench and recording the delivery volume at various pressures for several iterations of a particular model of gas valve.  Uwe, the head of the quality lab, was experimenting that summer with a small pump similar to the ULKA type, a tuned mass oscillator with a pair of check valves - one each at both inlet and outlet.  The spring and mass were tuned to resonate at about 55Hz - halfway between the North American 60Hz and the European 50Hz to see how well a single design pump could be made to work for both markets.  A solenoid coil was wrapped around the outside of the housing.  As I recall, the max pressure was something like 12 or 15 bar (175-200psi, approximately), but at an extremely low delivery rate - almost nothing.  At lower pressures, however, the flow rate was quite impressive, for a little gizmo that just sat there and buzzed.  We were experimenting using some sort of light oil (maybe 10 weight or ATF); I'm sure the viscosity of the working fluid would have also affected the resonant frequency and performance (added damping and losses as viscosity increased).

Anyway, we (well, mostly Uwe, while I watched and marvelled!) definitely used alternating current to drive the little experimental pump.