Electric water (coolant) pump(s)

[Christopher Schimke]

While I do have a thread going specifically about my 1.8T AEB installation, this latest addition may be useful to others who may other conversions, hence the new thread.

So here is series of events and observations that led up to the subject of this thread:

- When the engine was first installed it was summer time and I noticed that the engine ran pretty cool (170ºF or so on the ScanGauge – the coolant temp sensor that the ScanGauge reads is right off the back of the head where the coolant exits the engine and is the hottest). Another person with an AEB conversion also noticed relatively cool temps, so I didn't worry too much about it. At that time, I did not have the correct temp sensor installed to make the gauge on the dash operate correctly...or so I thought.

- As the weather started getting colder, I noticed that it took the engine a really long time to warm up. I had previously removed the ScanGauge, so from that point on, I didn't have any true indication of the engine temps (remember that the gauge temp sensor was wrong). The long warm up times were indicated by the lack of heat in the cabin. It was assumed that the thermostat was bad, so I installed a new one from Napa. At the same time, I reinstalled the ScanGauge and installed the proper (new) gauge temp sensor.

- Observing (on the ScanGauge) the typical coolant warm up routine (a mix of relatively short durations of non-stop driving combined with plenty of stop and go in traffic), the new thermostat was not opening until the ScanGauge read 272F with no heat in the cabin until that time. At the same time, with the new coolant gauge sensor in place, the coolant gauge would mimic the observations shown in the ScanGauge. The interesting part is that the new coolant sensor acted just like the old sensor, which I assumed was either the incorrect part or faulty. Turns out that the original sensor was correct and the engine was actually overheating. Assuming the new Napa, thermostat was bad, another new thermostat was installed with the same results. Thinking that the aftermarket thermostats were not performing correctly, a new OE Audi thermostat was installed with the exact same results. It was discovered that if I drove without stopping for 10-15 minutes, the thermostat would open at around 262F and the temps would drop down to 195F and hold there for as long as I drove. The heat in the cabin would begin when the thermostat opened. The moment I stopped, the coolant temps would start to rise.

- Looking at the Audi coolant schematics, it shows that the heater circuit flows from the back of the head, through the heater core(s) and feeds directly to the engine side of the thermostat. Thinking that maybe my front heater core (original) was plugged, thus preventing the warmed coolant from feeding the thermostat), I installed a brand new heater core. The results were better in that the heat would enter the core much sooner than before. However, the coolant was still getting up to 272F before the thermostat opened and the heater would go cold with extended idling. In addition to the coolant going cold in the heater circuit at idle, the engine temps would quickly rise from 195F up to a max of 272F. Once back under way (engine RPM increased), the coolant in the heater circuit would once again flow warm and the temps on the ScanGauge would drop to normal.

- Thinking that the increased friction involved in the new cooling system (compared to the original Audi application) was too much for the water pump at idle, a booster pump was added to the heater circuit. The intent behind this was not only to keep the coolant flowing through the heater circuit for heat in the cabin, but also to keep the heated coolant flowing to the backside of the thermostat in hopes that it would stay open at idle when the coolant flow was low.

- The addition of the booster pump helped the heat in the cabin tremendously, but it was not perfect. In stop and go traffic, warm coolant would flow to the heater core much more consistently than before, but the coolant temps observed on the ScanGauge were still showing that the thermostat was opening with the coolant in the engine being much too hot, but would stay open as long as the engine RPM was high enough to keep the coolant flowing. Once the engine RPM dropped to idle (and even just above idle, such as slowly rolling with traffic at very low RPM), the thermostat would close causing the observed coolant temps to spike.

- An inspection showed that the water pump was in good working order (no broken impeller, tight to the shaft, replaced with new anyway). It was also observed that the feed line to the radiator was usually only luke warm and the return line from the radiator (brand new radiator) was very cold, even when the observed engine temps were very high. Thinking that the coolant was possibly bypassing the main coolant lines through the heater circuit, an inspection proved this not to be the case. In addition, a complete reinspection and bleeding of the cooling system showed that everything was connected and bled properly and followed the original Audi coolant routes exactly. In other words, nothing is wrong the routing or bleeding of any of the coolant lines.

- After many conversations with a few Audi techs and lots and lots of reading about cooling systems, the only conclusion that I could come up with is that the original Audi water pump does not create enough head (pump outlet pressure) to keep the coolant flowing with the increased pressure drop (backpressure) in the Vanagon coolant system compared to the original Audi cooling system. The much longer coolant lines that connect the engine to the radiator and heater core create more friction for the coolant the have to push against as it flows through the circuits. Assuming the Audi pump does not create enough head for these conditions when the engine RPM (and therefore the water pump RPM) are low, the flow of coolant will drop dramatically or even stop altogether. While I don’t have a good way to test the actual pressure at the water pump outlet, all of the evidence points to this fact. This theory was somewhat validated/substantiated by another 1.8T owner here on the Samba who had to install a booster pump in his heater circuit to keep the temps up at idle.

- So here is my theory:
With the Audi water pump not having enough head (pressure at the pump outlet) at lower RPM, the engine is basically overheating at idle due to lack of coolant flow. Once the coolant temps reach 272F, the thermostat finally opens. However, the coolant in the system is stone cold and there is so much coolant volume that the thermostat quickly closes back up. This cycle happens over and over again when the engine is at very low RPM. At higher engine RPM, the pump creates enough head to keep the coolant flowing fast enough that the coolant temps in the entire system stabilize which keeps the thermostat in some state of open. This allows the coolant to continue to flow through the entire system, thus keeping the coolant temps in check. Once the engine RPM drops, the water pump can’t flow enough coolant and the observed engine temps start to rise. Only when the engine RPM is increased does the coolant temp start to drop due to increased flow. To prove this theory, I removed the thermostat, with the thinking that the thermostat adds some restriction to the system, even when open, and that without that restriction, the pressure drop in the system is reduced and the pump should be able to flow more freely. Sure enough! With the thermostat removed, the system never once showed the high temps at idle. Granted, without the thermostat in place and the ambient temps in the low 40s, the overall coolant temps were way down, so temps through the heater circuit were low as well. I think that proves that the culprit is the increased pressure drop (backpressure) in the system compared to the stock Audi system and the water pump’s inability to cope with it.

Why this is a problem in my system, I have no idea. Since I have never heard of anyone else having this same issue with an AEB conversion, I can only assume that there is something odd with my system that is causing this. Having spent way too much time on it with very little to show for, I was desperate to find a cure. Since everything seemed to point at the water pump/increased friction in the system as being the culprit, I decided to make a drastic move and bypass the mechanical water pump altogether and install a digitally controlled electric water pump.

I chose to install a Davies Craig EWP115 with their EWP Digital Controller (Kit #8030). The pump puts out 115 litres per minute and has a claimed 7000 hour lifespan when used with the EWP Digital Controller. The pump controller is pretty cool. Combined with the supplied thermal sensor, the controller’s main functions are to control the coolant flow via controlling the pump’s speed as needed based on the coolant temps and to run the pump after the engine shuts down to help prevent heat soak in the engine. You can set the required running temperature for your engine via the controller.

Here is Davies Craig’s explanation of how the controller works:

[insyncro]

Very cool (hot actually) Christopher

Every van would bleed easily without a thermostat.

I have experience with doing roughly the same with engines in race vehicles to relieve the engine of additional stress and robbed HP from belts and peripherals.
I swapped a 350 Chevy into my Bro's 84 Toyota pickup truck and have everything running electrically powered.
Man that sucker develops some HP.

I can think of a few different vanagon engine swaps that would benefit from this.

Thanks for sharing.

[seanjenn]

Feaseable for the WBXr engines?
Seems like replacing a waterpump these days is a dime toss quality-wise.
_________________
1987 GL Sunroof
2.1 4 spd

[insyncro]

[xoo00oox]

Hi Chris, I'm thinking about your original cooling problem. For those who don't know, I have the identical engine conversion except mine is at 50* and his is at 15*. My temp gauge has always read a little low but very steady even now driving with temps in the 20's. Heat is good, maybe just a bit cooler than with a stock 2.1. I have no extra electric coolant pump, just ran the two radiator hoses and 2 heater hoses where they attached to the engine as if it were in the Audi or Passat like you originally had done. Why would mine be fine and yours acting up?
I'm looking back at you build thread on page 4 where you changed the belt set up. It looks like the coolant pump is now running faster than stock, could this be your problem?

Original belt routing....

Your modified belt routing....

[Christopher Schimke]

[xoo00oox]

One other difference between our vans is the diameter of the long coolant pipes. I have the 2.1 plastic ones and assume you have the smaller metal lines, correct.
Possibly too small? I would not have thought so but maybe?

[insyncro]

[Christopher Schimke]

[Christopher Schimke]

[Crankey]

cool beans man !
I have heater issues too but for different reasons I think. I wondered about an electric pump...but I will deal with the problem I think I have first. will post up my own solutions when I get going on them.

the longer runs of coolant lines in our vehicles do seem to factor in when using engines designed for the front. all part of the fun I spose

[xoo00oox]

The stock 1.6na diesel engine van uses the same big radiator and uses the same coolant pump as a rabbit does and has no problem pushing coolant through the long lines, I would think your engine would be fine with these as well.
This is very strange to me as to why yours does this and mine does not. Can you think of any other differences between our two other than what we already mentioned? Is there any possibility that your pump running faster than stock could cause this?

[insyncro]

[Christopher Schimke]

[J Charlton]

I remember when I did my 2.5 suby conversion I had the tremendous benefit of having Tom Sheils reasonably near by. He came over at initial "fire up" time and worked with me to ensure that all was well and working fine. At the time, I remember asking him what the purpose of his thermostat adapter was. He replied that (as best I can remember) that the suby engine was very efficient in terms of not producing great amounts of heat and that the vanagon cooling system was very large relative to that in any of the subaru vehicles that the engines would have come out of .... this if everything was plumbed as regular in a donor vehicle, would result in the thermostat being in a constant state of opening and closing.
Thermostat bypasses are pretty regularly used on vehicles to provide quick heat to heaters and defrosters, long before the entire cooling system is up to temp and they do so quite effectively.
The effectiveness of Tom's bypass thermostat housing shown here (from the website http://www.subaruvanagon.com/tom/Thermostat%20housingk.htm

is that engine temperature coolant is "dropped" onto the working parts of the thermostat - i.e. the part with the bimetallic spring that expands forcing the thermostat open. This means that a small amount of coolant from the rad loop and heater loop is constantly being let into the engine cooling circuit once the engine itself is up to operating temperature. The amount let through the thermostat from the rad cooling loop increases as the entire system gradually warms up and the thermostat opens. The advantage of this is that it is gradual, not continually open and closed - without the bypass, the thermostat opens only to be subjected to a sudden influx of cold coolant, causing it to close, meanwhile the coolant in the engine block is already hot - leading to a feast or famine scenario of overheating and rapid closing, even partially of the thermostat due to the cooler coolant from the rad coming to it. Tom's bypass ensures that the thermostat is opening or closing in response to the temperature of the coolant coming from the engine only. You can see in the picture above that the input coolant barb is going to supplying the thermostat with warm/hot coolant on its "working part"
I've often thought that such a bypass would be a really nice piece to have for other engine conversions as well - especially TDI conversions.
I'm sure that if my memory or logic is faulty here that others will correct me.
_________________
NAHT hightop availability May 18 2023 -
Bend Oregon - for Oregon, California- (7 tot , 3 available), Kennewick Wa (6 tot, 1 available), Small Car Performance Fife Wa. (7 tot 4 avail ), Fairbanks Alaska (1 tot 0 avail)
Future availability TBD : Springfield Mass. Staunton Va, Florida, Colorado, Grand Junction Co., SLC probably late 2024

[WLD*WSTY]

It appears the EWP 80 pump would be an excellent choice as well. It is rated for N/A or turbo engines up to 3.0 liters, and pumps 80 lpm instead of 115 lpm. It also can be used in tandem with the mechanical pump, so providing redundancy should the electric pump pack it in.
_________________
'82 SyncroWesty, the first ever conversion.'06 Subaru 2.5L

[Christopher Schimke]

John, I considered doing something very similar to Tom Shiels' Subaru bypass system. I do think that it could be beneficial. To be honest though, I was tired of experimenting with things and just needed something that would work. My concern with my particular system and a Tom Shiels-like bypass is that I was (and still am not) convinced that the water pump has enough head to cover the increased friction in the system. Creating a bypass system like that would take time with no guarantee that it would solve all of my problems. I may take the time to build one in the future when I have more time and the weather isn't so bad, if for nothing else to prove or disprove the point.

*EDIT*

Audi does have a circuit system in place that feeds hot coolant into the inlet hose right before the thermostat with a reasonably substantial amount of hot coolant from the pressure tank (3/4" hose from the pressure tank feeding directly into the inlet hose). The difference is that this hot coolant enters the inlet hose about a foot away from the thermostat on the inlet side rather than right at the thermostat's outlet side like in the Tom Shiels system. Maybe with the extremely cold coolant coming in from the radiator, the influx of hot coolant from the pressure tank still isn't enough to prevent the thermostat from closing? I'm not sure.

[J Charlton]

I should have added in the post above that I really like the idea of an electric water pump and I've been sitting here all evening sketching and thinking how nice it would be to have a system driven by an electric coolant pump and controlled by a series of solenoid valves that would respond to temperature sensors. As far as failure of the system on the road - it would probably be quicker to replace the coolant pump with another than it would be to replace a belt driven mechanical pump. maybe a bit more costly to carry a spare though. I really like the concept of the coolant pump continuing to circulate coolant for a period of time after ignition off - gotta be better for any engine.
While I'm at it - I'd really like to fool around with a dry sump oiling system as well.... I think that it would also have a lot of very positive benefits.
_________________
NAHT hightop availability May 18 2023 -
Bend Oregon - for Oregon, California- (7 tot , 3 available), Kennewick Wa (6 tot, 1 available), Small Car Performance Fife Wa. (7 tot 4 avail ), Fairbanks Alaska (1 tot 0 avail)
Future availability TBD : Springfield Mass. Staunton Va, Florida, Colorado, Grand Junction Co., SLC probably late 2024

[Alaric.H]

I have been running a electric pump in my car for years with no problems.

[J Charlton]

Not to beat this thread to death, but, I had a pm questioning why the vanagon cooling system seemed to be appropriately sized and worked well with the wbx engines so why not with other replacement engines?
We have to get into the Law of Conservation of Energy here for a real explanation. The Law of Conservation of Energy states that energy cannot be created or destroyed, its form can change around, but in total, energy in = energy out.
In a van, driving down the road, the energy in a quantity of fuel is changed into kinetic energy (motion), thermal energy (heat) and sound.
If we ignore the sound portion for now and focus on the other 2, then
if we have 2 identical vans, one with a wbx engine getting eg 18 mpg, the other with a suby (or Zetech or 1.8T or whatever) getting (for the ease of colculations 24mpg. That means, that for the wbx engine to push its van down the road the same distance as the "other engined" van it takes 1/3 more fuel. If the vans have travelled the same distance at the same speed then where has the extra fuel gone? It was not used to create kinetic energy so it must have produced (primarily)heat.
How much heat? Someone can chime in here with the the BTU or caloric or # of joules per litre of gasoline - but to get a handle on how much heat, 1/3 of a gallon of gasoline is roughly 1 1/4 litres. If you were to ignite 1 1/4 litres of gasoline, how much heat would you get? - I don't know exactly, but its a lot.

When we talk about "fuel efficiency" we refer to the relative amount of fuel required to produce the kinetic energy that we want (motion). More fuel efficient power plants will therefore produce more kinetic energy per unit of fuel that will less efficient ones. They will also produce less "waste" energy (heat). If the vanagon coolant system was properly sized for the wbx engine, then it will be relatively oversized for a more fuel efficient engine - being oversized is not necessarily problematic inasmuch as the primary function of the cooling system is to keep the engine from overheating, yet when we add complexity to the system to reclaim (use) some of the waste energy (eg heater/defrosters) and some form of controlling device to maintain proper engine operating temperatures (thermostat) we have to take into account the effect of the reduced thermal loading of the system on the controlling device.
So - back to the thermostat adapter - we need such a device for other conversion engines to better manage the effect of the cooling system when it is at the point of the thermostat opening and closing throughout warm up until the system is completely warm. In colder climates, that opening closing cycle may well go on for quite a while as the vanagon rad is quite capable of removing enough heat from the coolant to have the coolant returning to the engine to be at sub thermostat opening temperatures.
Some folks who have switched TD diesel engines to TDI engines have commented on the increased time it takes to "warm up".
_________________
NAHT hightop availability May 18 2023 -
Bend Oregon - for Oregon, California- (7 tot , 3 available), Kennewick Wa (6 tot, 1 available), Small Car Performance Fife Wa. (7 tot 4 avail ), Fairbanks Alaska (1 tot 0 avail)
Future availability TBD : Springfield Mass. Staunton Va, Florida, Colorado, Grand Junction Co., SLC probably late 2024