Electric water pumps vs mechanical
Discussion
Search is down again, so I can't find the previous discussions. I thought about an electric water pump for my TVR to solve a problem, but I fear it may cause at least 5 more... Here's my scenario based on some very crude first order approximations:
The engine produces a maximum of 240kw of heat (based on 1/3 of total energy in petrol going in to cooling water)
There's about 6 litres of water inside the engine at any particular time
The specific heat capacity of water is 4.181 J/gk
So, based on 1W = 1J/s I can say that d(temperature)/dt is approximately 240/(6000*4.181) = 9.5 degrees/sec at full power.
An average electric pump flows 115 litres/hour or 2 litres/sec. This to my mind looks like the temperature rise of the water will become close to 30 degrees/sec - rather a toasty engine! The radiator is unlikely to deal with this temperature - best guess is that it will drop the water temperature by 15C or so. At least the pump only consumes 150W while allowing the engine to boil...
Given the spec on the pump says it can deal with 400bhp, where is the mistake in my logic? I seem to be off by a factor of 2 if the spec is correct and not based on a very low duty cycle of operation.
As a comparison, does anyone know the power consumption and flow rate of a belt driven water pump (Rover v8 for preference)? I appreciate it all depends on RPM, so let's say at a peak power of 5800 RPM to keep it simple.
Thanks.
The engine produces a maximum of 240kw of heat (based on 1/3 of total energy in petrol going in to cooling water)
There's about 6 litres of water inside the engine at any particular time
The specific heat capacity of water is 4.181 J/gk
So, based on 1W = 1J/s I can say that d(temperature)/dt is approximately 240/(6000*4.181) = 9.5 degrees/sec at full power.
An average electric pump flows 115 litres/hour or 2 litres/sec. This to my mind looks like the temperature rise of the water will become close to 30 degrees/sec - rather a toasty engine! The radiator is unlikely to deal with this temperature - best guess is that it will drop the water temperature by 15C or so. At least the pump only consumes 150W while allowing the engine to boil...
Given the spec on the pump says it can deal with 400bhp, where is the mistake in my logic? I seem to be off by a factor of 2 if the spec is correct and not based on a very low duty cycle of operation.
As a comparison, does anyone know the power consumption and flow rate of a belt driven water pump (Rover v8 for preference)? I appreciate it all depends on RPM, so let's say at a peak power of 5800 RPM to keep it simple.
Thanks.
I tried one of the s
tty plastic Davies Craig pumps a few years back on my RV8.
I deemed it to be rubbish. As you say, awkward to fit and in my case the engine struggled with temperatures even when the pump was flat out.
The standard pump is already there, it works and is very reliable.
I'd struggle to find any reason to want to change that !

I deemed it to be rubbish. As you say, awkward to fit and in my case the engine struggled with temperatures even when the pump was flat out.
The standard pump is already there, it works and is very reliable.
I'd struggle to find any reason to want to change that !
There are only 2 or 3 electric pumps that have a sufficient flow rate at the required head to generate a suitable DeltaT across your engine (you'd be aiming for ~6degC) to completely replace the mech pump.
1st one in the Pierburb CWA 200 OEM brushless water pump:
http://www.kspg-ag.de/pdfdoc/kspg_produktbroschuer...
but it needs CAN to command it, you can't just plug it in at it works.
2nd one is the Meziere 55gpm remote mounted pump
http://www.meziere.com/ps-1180-1148-wp365s.aspx
There is also an OEM style 12v brushless one, but i can't currently remember who makes it, but they are used by the "supercar" OEM's as booster pumps quite a bit.
DO NOT UNDER ANY CIRCUMSTANCES CONSIDER USING ANY OF THE DAVIESCRAIG EWP UNITS, UNLESS YOU WANT TO MELT YOU ENGINE!
Generally speaking, overheating engines are a result of insufficient heat rejection from the radiator rather than a limitation in system flow capability at low engine rpm
1st one in the Pierburb CWA 200 OEM brushless water pump:
http://www.kspg-ag.de/pdfdoc/kspg_produktbroschuer...
but it needs CAN to command it, you can't just plug it in at it works.
2nd one is the Meziere 55gpm remote mounted pump
http://www.meziere.com/ps-1180-1148-wp365s.aspx
There is also an OEM style 12v brushless one, but i can't currently remember who makes it, but they are used by the "supercar" OEM's as booster pumps quite a bit.
DO NOT UNDER ANY CIRCUMSTANCES CONSIDER USING ANY OF THE DAVIESCRAIG EWP UNITS, UNLESS YOU WANT TO MELT YOU ENGINE!
Generally speaking, overheating engines are a result of insufficient heat rejection from the radiator rather than a limitation in system flow capability at low engine rpm
Max, what about the Stewar Components stuff. They are very nice little pumps and they flow good amounts of water.
I personally have no interest, but I did see Prodrive using them on their LeMans engines.
http://www.stewartcomponents.net/Merchant2/merchan...
I personally have no interest, but I did see Prodrive using them on their LeMans engines.
http://www.stewartcomponents.net/Merchant2/merchan...
The pumps i couldn't remember the name of are:
http://www.emp-corp.com/products/advanced/WP29-ele...
The Stewart pumnps are ok, when we did the DBRS9 program we spent ages trying to find a suitable pump, and the choice was much much more limiting back then. The CraigDavid (alright Kes) pumps don't flow at all at >1bar head, and the Stewart pump was the only other option tbh.
Sat here in 2012, i'd use either the OEM Pierburg with a CAN gateway module to control it, or the EMP pump shown above.
http://www.emp-corp.com/products/advanced/WP29-ele...
The Stewart pumnps are ok, when we did the DBRS9 program we spent ages trying to find a suitable pump, and the choice was much much more limiting back then. The CraigDavid (alright Kes) pumps don't flow at all at >1bar head, and the Stewart pump was the only other option tbh.
Sat here in 2012, i'd use either the OEM Pierburg with a CAN gateway module to control it, or the EMP pump shown above.
Max_Torque said:
I have used one of these and it's an excellent unit. Extremely well made and has the guts behind it too.One word of caution: Meziere pumps don't tolerate dry running AT ALL (the bearings don't like it), so follow the bleeding process religiously. There's a bleed screw on the pump body.
Hi all! If I want to use one of the three EWP you mentioned above should I retain the mechanical pump and use electrical as auxiliary or can I use electrical alone? I want to instal a EWP into my Porsche and gain a few HP and also hava less heat in the engine so I gain in less consum of fuel, longetivity and emisions, Im right?
Seems to me like nobody has addressed this particular scenario, rather the all or nothing EWP.
If you want to rev the car more than it was originally intended for (ie OP Land Rover Pump in a TVR!) the pump needs regearing with a smaller pump pulley (the larger crank pulley is not possible in OP's case without reducing the pump pulley as they already nearly kiss). That leaves the pump probably under spec at low rpm when it would seem assistance from an electric pump might be attractive.
Pity you can't have something like a viscous coupling that varies the pump rotation not exactly with engine rpm?
If you want to rev the car more than it was originally intended for (ie OP Land Rover Pump in a TVR!) the pump needs regearing with a smaller pump pulley (the larger crank pulley is not possible in OP's case without reducing the pump pulley as they already nearly kiss). That leaves the pump probably under spec at low rpm when it would seem assistance from an electric pump might be attractive.
Pity you can't have something like a viscous coupling that varies the pump rotation not exactly with engine rpm?
on the electric pumps... i've not been able to fully work out how the Mezerie pumps control the flow.... or do they run flat out all the time?
Id like to loose the themostat on my engine as I believe its restrictive... they make a new one with a built in stat that I've been looking at.
I rev my car nearly 2k more than the stock pump was orignaly designed for... it copes on most other cars but Ive personally destroyed 2 that have come appart inside and it cant be conicidence as I do thrash mine to the limiter on track.
Id like to loose the themostat on my engine as I believe its restrictive... they make a new one with a built in stat that I've been looking at.
I rev my car nearly 2k more than the stock pump was orignaly designed for... it copes on most other cars but Ive personally destroyed 2 that have come appart inside and it cant be conicidence as I do thrash mine to the limiter on track.
I'd say it would be up to you to sort out pump control. The Meziere's etc probably do just run flat out
The Stewart Components stuff do have some sort of controller on them, not sure how it's used though.
Or you could maybe use your ecu, configure a PWM output to control the pump and use a solid state relay to handle the current.
The Stewart Components stuff do have some sort of controller on them, not sure how it's used though.
Or you could maybe use your ecu, configure a PWM output to control the pump and use a solid state relay to handle the current.
to me it seems that these PWM are complex and hard to get right... certainly the one of the people I consider experts on here (max) started a fuel pump one which never got finished hence my concern Id never get it done right 
This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?

This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?
andygtt said:
to me it seems that these PWM are complex and hard to get right... certainly the one of the people I consider experts on here (max) started a fuel pump one which never got finished hence my concern Id never get it done right 
This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?
Max did finish the fuel pump controller, and did use it on his car from what I recall.
This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?
But if you're wanting to switch mechanically driven components to electrical, then of course you need to consider what powers the electrics. The same drive belt that once powered the water pump lol.
stevieturbo said:
andygtt said:
to me it seems that these PWM are complex and hard to get right... certainly the one of the people I consider experts on here (max) started a fuel pump one which never got finished hence my concern Id never get it done right 
This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?
Max did finish the fuel pump controller, and did use it on his car from what I recall.
This is the route Id like to take... not sure I have enough outputs left in my Motec to do it though. I already have 2 for the fuel pumps (I switch one on and then use another switch to turn the other one on at set parameters) and 1 for the gearbox oil pump.
If the water pump ran flat out all the time, wont it put a real strain on my battery on idle?
But if you're wanting to switch mechanically driven components to electrical, then of course you need to consider what powers the electrics. The same drive belt that once powered the water pump lol.
Also, with a conventional claw pole alternator only being approx ~65% efficient, it's generally better to use direct mechanical power rather than electromechanical power for "continuous loads"
andygtt said:
I rev my car nearly 2k more than the stock pump was orignaly designed for... it copes on most other cars but Ive personally destroyed 2 that have come appart inside and it cant be conicidence as I do thrash mine to the limiter on track.
Out of interest. Is it OEM pumps failing, or aftermarket ? And can you not just alter some pulley sizes to reduce pump speed ?or...as some drag racers do, they hook up an external motor which then drives the OEM pump instead. Not sure what sort of life span they have though, or how effective they are.
They are stock ford pumps, knowone else has had issues.... whats happening to mine is that the pump is alloy and has curved anti cavitate blades... in the houseing has a steel insert that the blades run in, its a thin sheet pressed in... this comes loose, as its just me and I do rev mine harder than others (ie right to redline theirs all have lower down powerbands so pointless hanging onto the revs on track, mine is different due to the huge single turbo) so must be the revs.
Being frank, the stock pump is designed to shed heat from an NA ford engine making under 250bhp.... Im running a huge amount more than that AND i'm mid engined... so Im thinking the stock pump may be marginal for my car anyhow.
So back to the PWM... anyone got a recommendation as to a reliable one that I could use? or should I just run the Mezerie one at full power all the time?
Im actually reconfiguring the battery etc ATM, I bought 2 lithium batteries that weigh around 1.6kgs each yet provide 500CCA (stock battery has 540CCA and is 15kgs)... I was planning to run one battery dedicated for the engine, pumps and ecu which switches off when ignition off... and the other for all the other electrics, thus if the lights or alarm flatten the battery it will still start.....
Im just worried Im going to give myself an issue on idle as I already run 4 fuel pumps (2 044's but only one runs on idle) plus an optional gearbox pump... now maybe a water pump.... going to become an all electric car soon lol
Being frank, the stock pump is designed to shed heat from an NA ford engine making under 250bhp.... Im running a huge amount more than that AND i'm mid engined... so Im thinking the stock pump may be marginal for my car anyhow.
So back to the PWM... anyone got a recommendation as to a reliable one that I could use? or should I just run the Mezerie one at full power all the time?
Im actually reconfiguring the battery etc ATM, I bought 2 lithium batteries that weigh around 1.6kgs each yet provide 500CCA (stock battery has 540CCA and is 15kgs)... I was planning to run one battery dedicated for the engine, pumps and ecu which switches off when ignition off... and the other for all the other electrics, thus if the lights or alarm flatten the battery it will still start.....
Im just worried Im going to give myself an issue on idle as I already run 4 fuel pumps (2 044's but only one runs on idle) plus an optional gearbox pump... now maybe a water pump.... going to become an all electric car soon lol
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff