Tracking Down Coolant Leaks

This forum is specifically for engine and gearbox posts, please try to keep on topic in any thread to make things easy to find

Moderator: Board Moderators

Post Reply
User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Tracking Down Coolant Leaks

Post by rbgosling »

OK, in the last month or so I've had more practice at this than I would like, so I thought it would be worth sharing what I've been doing.

Undoubtedly the most useful bit of kit to help with this is a coolant pressure test kit. They can be picked up pretty cheaply from eBay, but you are welcome to borrow mine for the cost of postage - once I've fixed all my leaks!

Image

I drained all the coolant out of the system when checking, on the basis that air escaping from a tiny hole is far easier to detect than water.

Pump the pump until you can hear some hissing somewhere, or until it's a couple of psi below the radiator cap pressure, whichever happens first.

Image

It's not as easy as you might think to determine exactly where the hissing is coming from, but a bit of silicone tube stuck in your ear as a "stethoscope" really helps:

Image Image

Once you have a suspect, a 50/50 mix of washing-up liquid and water can help to confirm. A small syringe with a short bit of hose is ideal to apply it to the suspect area, but in this case I just used a teaspoon:

See this video.

If you can pressurise the system and not hear any obvious hissing, still go around with the stethoscope and check all the ends of each coolant hose.

If you still don't find anything, leave the system pressurised for 5-10 minutes. If the pressure doesn't drop, you're good.

With my test kit, once above 10 psi or so it did have a tendency to leak at the tester cap, but a bit of wiggling usually managed to stop it.

If the pressure does still drop, then you must still have a leak somewhere you haven't identified. Obviously the heater matrix is hard to access and check, so don't go there until you've ruled everything else out (I went there too early and found it bone dry!). The only other option I can think of (please advise if there's anything else) is the head gasket, which is obviously more of a PITA to fix. But if there's a significant leak there, you'd expect other symptoms - either a white steamy exhaust (leak to the combustion chamber), mayonnaise in the oil system (leak between oil and coolant passages), or oil in the coolant (ditto).

Please do chime in with any useful experience, advice and/or tips anyone else has!
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

User avatar
bash
Senior Poster
Posts: 2222
Joined: Thu Nov 03, 2005 15:34
Model: Se, V8
Colour: white
Year: 1986
Location: Doncaster
Contact:

Re: Tracking Down Coolant Leaks

Post by bash »

The one hard to see place that sometimes leaks are the water passages in the inlet manifold, theres alot of 'stuff' hanging off some small studs (inlet, carbs, air filter). Perhaps while drained it might be worth a new gasket and sealant around the water passages. Just a thought.

Bash
Two things are infinite: the universe and human stupidity; and I'm not sure about the universe.

TonyL
Junior Poster
Posts: 81
Joined: Sat Dec 31, 2022 18:56
Model: Excel/Elite project
Colour: Red/Red (for now)
Location: Chobham

Re: Tracking Down Coolant Leaks

Post by TonyL »

Thanks for explaining all that, Richard.

Your stethoscope picture made me laugh.

Have you fixed all your leaks now, or is this still WIP?

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

Still WIP.

Yesterday I found a split in the short fat hose that goes from the water pump to the inlet manifold. This is new, it wasn't there when I did most of this work over the New Year, and it was losing coolant rapidly (see FaceBook for a pic!). I've temporarily stuck a bicycle tyre patch over the crack, which seems to be working fine, until I can replace the hose.

I also found a leak between the water pump and the block. In the post-85 engine there is no gasket here, just sealant. When I rebuilt the engine one of the bolts snapped, so it's now held with 4 bolts out of 5, and of course the leak is where the 5th bolt should be - it did manage to hold for 30k miles / 6 years! It's quite possibly only sprung a leak here because I've been testing it at higher pressures than it normally sees. So I'm going to have to pull the water pump off, pray that I can extract the remains of the broken bolt, and re-assemble with new sealant (currently on order).

Once the water pump to block sealant arrives, and some new bolts, I'll be able to address both these leaks. Then I test again, and see if I can provoke any more leaks...
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

User avatar
Lotus-e-Clan
Senior Poster
Posts: 4461
Joined: Thu Nov 30, 2006 13:25
Model: Excel SE - EWP/Waterless!
Colour: Not Blue or Green
Year: 1989
Location: Swaledale

Re: Tracking Down Coolant Leaks

Post by Lotus-e-Clan »

I like the tiny pipe stethoscope. A mini-me version of Ian Tyrrell's carb tuning tool!
I found a reference to the OE cap pressure value in the Excel SN.
There are two given. 10 psi or 15 psi, but no qualification as to application.
Which left me wondering if 15 psi is meant for export to hot climes as there's more need to raise the boiling point, and it's in the same section where the different climate fan switch value are given.
I can't think of another reason for using the higher pressure, except for use in hot climes (unless racing, perhaps)?

That said, if leak free was OE achievable at 15 psi in a 912 engine, then there's no reason why you shouldn't strive for it. But given the current age of the components in the cooling system, I'd have been happy, for now, to get it leak free at 10 psi for use in our temperate clime.
Peter K

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

According to the parts manual, it says "R/B 15 psi cap A082K6031F", implying that whatever isn't R/B should have the 10 psi cap A075K6004Z. But I'm not sure what the "R/B" stands for?

The SJ Sportscar parts page doesn't really make it much clearer. It does say that the 15 psi cap is for the early single cap tank. None of the part numbers on that page agree with the part numbers from the Lotus parts manual. And when you search for the 10 psi cap by part number, SJs re-direct you to a 7 psi cap.

Lotusbits list both 10 and 15 psi caps without saying anything about application.

Yes, I tend to agree, at this stage I'd be happy if it held 10 psi! Particularly as I am strongly considering swapping to Evans Waterless Coolant, once I've eradicated the leaks, and that should put very little pressure onto the system.
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

Zag
Regular Poster
Posts: 100
Joined: Mon Aug 28, 2017 04:35
Model: Excel SE
Colour: Ice Blue
Year: 1986
Location: East Yorkshire

Re: Tracking Down Coolant Leaks

Post by Zag »

R/B is Replaced By..

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

Zag wrote:
Thu Feb 01, 2024 09:42
R/B is Replaced By..
OK. Which would imply that they started with the 10 psi cap, but then Lotus in their infinite wisdom, for some reason that I'm sure is a good one but we don't know what, switched to 15 psi. Not wanting to second-guess their excellent engineers, I'll trust that there was a good reason for this and stick with what they did, and leave the 15 psi cap on my car.
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

My original header tank (1990 SE car) had a larger pressure cap with no ears, like this one from Lotusbits.

Image

Unfortunately there is no indication what the pressure limit on this cap is.
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

User avatar
Lotus-e-Clan
Senior Poster
Posts: 4461
Joined: Thu Nov 30, 2006 13:25
Model: Excel SE - EWP/Waterless!
Colour: Not Blue or Green
Year: 1989
Location: Swaledale

Re: Tracking Down Coolant Leaks

Post by Lotus-e-Clan »

You can test the cap pressure with your kit. Just raise the pressure until the seal lifts and you hear air from the overflow outlet on the tank.
I did this diy style prior to Evans conversion with a bike tyre valve pushed into a plastic wine cork pushed/clamped into a 5/8 (16 mm) hose attached to No 4 heater hose outlet. Used a track pump (gauged) to raise pressure above cap value. 👍
Peter K

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

Just been doing some Computational Fluid Dynamics on the Mercedes F1 engine. You really don't want to know what sort of coolant pressures that engine uses :shock:
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

Pete Boole
Senior Poster
Posts: 3722
Joined: Tue Apr 04, 2006 21:28
Model: Elite
Colour: Monaco White
Year: 1974
Location: Nottingham

Re: Tracking Down Coolant Leaks

Post by Pete Boole »

rbgosling wrote:
Thu Feb 01, 2024 13:09
Just been doing some Computational Fluid Dynamics on the Mercedes F1 engine. You really don't want to know what sort of coolant pressures that engine uses :shock:
I do!! :D

Pete

User avatar
rbgosling
Senior Poster
Posts: 1720
Joined: Sun Jul 04, 2010 22:55
Model: Lotus Excel SE
Colour: Midnight Blue
Year: 1990
Location: Rugby

Re: Tracking Down Coolant Leaks

Post by rbgosling »

We’re running at temperatures at which water has no right still being a liquid… but it is.

(I’ll probably get in trouble if I say any more! Try asking me in person when we next see each other ;) ).
"Farmer" Richard

1990 Lotus Excel SE (Lilith)
2022 MG MG5 EV (not due to be a classic for quite a few years...)
2011 Nissan Leaf (Ragly - EV pioneer, must be due to be a classic one day)

Zaphod
Senior Poster
Posts: 243
Joined: Mon Aug 23, 2021 21:38
Model: Excel SE
Colour: Blue
Year: 1990
Location: Uffington

Re: Tracking Down Coolant Leaks

Post by Zaphod »

I have used a marginally different approach, fill the system with water/washing up liquid, run to temp (to get the system to operating pressure) the leak will normally have bubbles coming out of it, (unless it runs at silly temps) if not apparent let it all cool down and empty it then pressuring it with air, the residual coolant in the area that is leaking will bubble
Of all the things I have lost I think I miss my mind the most...

1952 MG TD
1957 MGA 1500
1959 Austin Healey Sprite
1975 MG Midget 1500
1990 Lotus Excel

Post Reply