Using a second cat pump as an energy recovery device on an RO unit.

[mckenzie.keith]

It looks like cat pumps are commonly used as one of the key components of small RO desalinators.

These systems pump water up to high pressure using the cat pump, and then some of the water goes through the membrane leaving salt behind. The fresh water passing through the membrane is no longer at high pressure. So there is no potential energy there. But the brine which is left behind is still at high pressure. One way to dispose of this is to let it pass through a restrictive valve. This dissipates the energy of the stream.

The clark pumps in Spectra watermakers don't operate this way, and they are able to avoid this waste of energy and are thus much more efficient in terms of gallons per kWh of energy used.

An idea I have been thinking about for a while is, why not take a conventional system with a cat pump, and add a second cat pump facing backwards and running in reverse to recover the energy? See this thread for example.

The second pump could be mechanically coupled to the first one but running at a slower speed (belts with different pulley sizes), or it could have a motor operating in regen and pushing electrical power back into the system. I think the belt coupling would be way more efficient as long as the speed ratio does not need adjusting.

It looks like the cat pumps do not have any internal gear reduction or anything. So I think this might work unless the inlet and discharge valves are one-way only, which they may well be.

I am sure people will say "why would you want to do this?" Well, you know, I am a curious type of person. I am just wondering if it is even remotely practical. Someone might pick up a cat pump from a pressure washer or something dirt cheap and use it for energy recovery. Spectra systems are quite expensive!

 

[El Borracho]

Can a CAT pump be driven by fluid flow? I would guess not…because simple check valves vs. cam timed valves…but I have never inspected one.

If so, then the energy recovery pump needs to be arranged to work at a flow to match the membrane rejection ratio. That is how the Clark operates. A belt would be good. A motor would then drive the belt to add the power lost to friction, and etc.

 

[mckenzie.keith]

Can a CAT pump be driven by fluid flow? I would guess not…because simple check valves vs. cam timed valves…but I have never inspected one.

Yeah maybe not. I can't tell from the exploded diagram. (I am an EE not an ME). It would make sense to use simple check valves. The inlet pressure in the datasheet is too low to connect to the brine. So it would have to be running backwards. The valves might prevent that. I wonder if they could be reversed?

If so, then the energy recovery pump needs to be arranged to work at a flow to match the membrane rejection ratio.

Exactly.

That is how the Clark operates. A belt would be good. A motor would then drive the belt to add the power lost to friction, and etc.

Exactly what I was thinking.

 

[Max Rockatansky]

I think you’ll find that cat pumps don’t run well in reverse

 

[Ajax]

I don't ask "Why would you want to do this?"

I ask "Why hasn't this already been done?" Answering that question usually tells me why an idea won't work.

 

[2airishuman]

An idea I have been thinking about for a while is, why not take a conventional system with a cat pump, and add a second cat pump facing backwards and running in reverse to recover the energy? [...] So I think this might work unless the inlet and discharge valves are one-way only, which they may well be.

I've had pumps like that apart. Yes, the inlet and discharge valves are one-way only.

The Clark pump works by having opposing pistons of different sizes driving each other which introduces few losses since there is no rotating shaft. I think it's a really innovative design, and hard to improve upon. I believe it is patented but the patent is likely to expire soon. Perhaps this will lead to other sources for similar pumps.

 

[El Borracho]

Just for background….the pump would not be run “in reverse”. The flow would be forced thru in the normal direction in hope of turning the shaft. Because only check valves are used the flow would simply progress thru both inlet and outlet check valves without moving the piston. Hopeless.

The Clark pump is simple and effective. The basic function is used widely in industry. Must pre-date Spectra’s use. Dive shops use a similar thing to top up SCUBA tanks: Intensifier?

I have done what was recently mentioned in another thread: Stripped my Spectra down to the essentials. The various sensors and automatic valves failed one-by-one. The Spectra control software is the devil’s work. Support was less than helpful. Their brackets and chassis are inconvenient. So now it is arranged in the absolute minimum configuration and works far better. Smaller, lighter, far more reliable — however does demand a bit more geeky manual manipulation of valves and hoses to maintain. 20 years in service.

 

[Diarmuid]

IME, scrounging amps seldom works out to advantage. Which is a shame, b/c in small supply/demand systems like sailboats, you really do hate to see free energy just going to waste. So setting aside the cat pump idea, we can view the challenge as a general microhydro system. Backpressure isn't a problem, as you note: the brine is a wastewater product and you aren't wrecking any deltas by slowing down its exit. Microhydro has successfully been employed by many cabin-dwellers and has certain advantages over solar & wind: if you live in the Oregon foothills, for instance, your little turby can be making 200W, 24/7/365.

The three controlling factors for hydro are head, flow, and time. You need at least two of the three for the system to not totally suck. Head is pressure: you have plenty of that on the RO pass-thru. Flow is volume: that's probably not so great, even assuming anywhere from 3-25 gallons of bypass for every gallon of fresh produced. In seawater, probably we are toward the high end: make 25 gallons of potable water, bypass 400 gallons. That seems like a lot, but in terms of hydrogeneration, it isn't really. A hydro system on quite a small mountain stream will exceed 400 gallons in an hour.

Then there's time: running your RO for two hours every couple of days defeats the slow-but-steady benefits of microhydro. The units aren't typically large or efficient, typically checking in at 200-300W peak. They are usually simple drag (turgo) impellers with heavy losses. Is it worth all the complexity & headaches for 0.3kWh/day? And there will be headaches. Especially in high-head systems, parts erosion occurs at a frightful pace -- even in clear fresh water. Salt water destroys everything much faster. Brine would act like a freaking waterjet cutter.

 

[Diarmuid]

Secondary but important question: do you always run your engine or generator while making water? (Or only make water while battery charging or motoring?)

 

[El Borracho]

Secondary but important question: do you always run your engine or generator while making water? (Or only make water while battery charging or motoring?)

I try to make water when the sun is shining. Only bcuz the charge efficiency of LA batteries is so poor.

If I were designing an RO system for small yachts I would try to create a micro system that ran all the time. Like 5 gallons a day. Saves the flushing waste. Might reduce energy, size, weight, noise, operator attention.

 

[221J]

A salinity differential battery could recoup some of the energy put into the RO. The brine is the high salinity source and seawater is the low salinity source. I have no idea what percentage of your hull volume the battery would consume. The pressure in the brine can supplant the pump.

 

[mckenzie.keith]

Just for background….the pump would not be run “in reverse”. The flow would be forced thru in the normal direction in hope of turning the shaft. Because only check valves are used the flow would simply progress thru both inlet and outlet check valves without moving the piston. Hopeless.

The Clark pump is simple and effective. The basic function is used widely in industry. Must pre-date Spectra’s use. Dive shops use a similar thing to top up SCUBA tanks: Intensifier?

I have done what was recently mentioned in another thread: Stripped my Spectra down to the essentials. The various sensors and automatic valves failed one-by-one. The Spectra control software is the devil’s work. Support was less than helpful. Their brackets and chassis are inconvenient. So now it is arranged in the absolute minimum configuration and works far better. Smaller, lighter, far more reliable — however does demand a bit more geeky manual manipulation of valves and hoses to maintain. 20 years in service.

In addition to the straight passthrough problem due to the simple one-way valves, the pumps (according to the datasheet) have a much lower pressure rating for the inlet. This is why I proposed hooking it up backwards. The idea was to reverse both flow and direction of rotation. Just explaining my rationale.