LPG or CNG: Why must we run Petrol at start-up?

Post Reply
Message
Author
CNG
Member
Posts: 260
Joined: Sat Oct 20, 2012 5:54 pm

LPG or CNG: Why must we run Petrol at start-up?

#1 Post by CNG » Wed Nov 24, 2021 8:19 pm

I'm very much used to my single-point and stepper motor systems being more than happy to start-up, and be running LPG/CNG from the off, this, even in freezing conditions. My stepper motor Necam system wasn't quite so impressive yet is usually seen running gas by 20 seconds at most - 5-6 seconds in summer. Now, one big advantage of gas is that it doesn't splash neat petrol around a cold engine. Thus, if there's one time we really want things to run gas, it's at start-up. Multi-point set-ups seem to struggle in this regard, why my Necam system with just as many ice-up points will breeze a cold-start, I don't understand.

My current multi-point Nevo ECU can take over a mile or 5-6 minutes if it's cold, I'm guessing it's to do with injectors icing? Maybe not, what is actually happening, what stops us running gas from the off?
Last edited by CNG on Wed Nov 24, 2021 8:31 pm, edited 1 time in total.

Brian_H
Intermediate Member
Posts: 2316
Joined: Mon Jan 23, 2012 12:21 pm
Location: Milton Keynes

Re: LPG or CNG: Why must we run Petrol at start-up?

#2 Post by Brian_H » Wed Nov 24, 2021 8:30 pm

The main reason its done that way, is to avoid over fueling at cold temps due to cold start enrichment on the petrol ecu. The LPG ecu can't tell if its active or not, and doesn't compensate for it. You can also find with injectors that openation on particually cold temps cause premature failure. You should be ok with a coolant temp switchover around 30 degrees usually, I guess the questions you need to answer is what is it set at, and does the temperature indicated in the software look somewhere near accurate? (the former is fairly simple to understand, the later could be wrong type of sensor selected, faulty sensor, or badly placed sensor depending what you have, or poor coolant flow for some other reason)

You may not have as much of an issue on CNG as LPG as its obviously not liquid coming to the front end which can cause problems with LPG - No idea if its an issue with CNG so can't advise there.

Generally mine switches over within a mile of driving (the first mile from home is mostly downhill with little revs needed and slow speeds) when its cold like the present time, shorter if its mid summer and warmer. Thats the same Nevo on 4 different engines though the LPG varient rather than CNG here. As you said the Necam I had would switch over generally before I'd gone to pull away alot of the time, but the Necam is different in the respect of a full map thats vehicle specific loaded onto it. Only reason that starts on petrol as far as I know was to avoid backfires from gas flooded manifold if it failed to start for some other reason.

There are other thresholds that can delay switch over, which may be set as well in the software, but I'd start with checking the sensor data and the basic switchover threshold first.

LPGC
Installer
Posts: 4454
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: LPG or CNG: Why must we run Petrol at start-up?

#3 Post by LPGC » Wed Nov 24, 2021 8:35 pm

Because of the choke effect - That splashing neat petrol around a cold engine means that an excess fuel factor is necessary to compensate for the petrol that doesn't atomise so won't form part of the effective fuel mixture, the choke makes up for what would otherwise be a lean mixture. Most CNG/LPG systems are slave systems (to the petrol system), so would react to the choke effect by injecting an excess of LPG, LPG is a perfectly atomised vapour so doesn't need the choke effect, so the choke effect would make the mixture too rich.

And because a cold pressure reducer will output liquid gas instead of vapour, the injectors need to be fed with vapour.

You can get away with starting an engine running an LPG mixer system on LPG because mixer systems are standalone and won't react to the petrol system's choke effect. And because a mixer system reducer outputs gas at atmospheric pressure (lower pressure than a sequential) system, the lower pressure means gas is more likely to vaporise at any given (cold) temperature.

Post crossed with Bri's.
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

CNG
Member
Posts: 260
Joined: Sat Oct 20, 2012 5:54 pm

Re: LPG or CNG: Why must we run Petrol at start-up?

#4 Post by CNG » Wed Nov 24, 2021 8:45 pm

Ah, got it, thanks, and nothing to do with icing-up. These systems, were they not sat on the petrol ECU 'could' run gas from the off, much as a Necam. As for my set-up and its Nevo, the project has come to halt, the engine has had it. Doing a rebuild on a spare as we speak. I'll get Dai 'classicswede' to set-up the delay once i can give him a decent workable engine, maybe CNG will be better at dealing with cold-starts. At least now I understand it all a bit more. Thanks.

Gilbertd
Advanced Member
Posts: 8748
Joined: Wed Apr 09, 2008 10:00 am
Location: Peterborough

Re: LPG or CNG: Why must we run Petrol at start-up?

#5 Post by Gilbertd » Wed Nov 24, 2021 9:13 pm

If you were to run a completely stand alone LPG or CNG system using something like a Megasquirt ECU, you could run an injection system from cold (as Robertxx has done on at least one of his cars). No idea about CNG, and can't be bothered to look it up, but the boiling point of LPG is -44 C so as long as the vaporiser temperature remains above that, you won't get icing, no matter how cold the ambient is. When I first bought my Range Rover Classic with a single point, the previous owner told me if he tried to run on gas from stone cold, it would die within 400 yards. I found that, sure enough, it would, as the vaporiser had been plumbed in parallel with the heater and the coolant was taking the path of least resistance, the heater matrix, and not keeping the vaporiser temperature above -44 C. Result was this:

Image

The solution was to change the plumbing to series so the coolant flow went through the vaporiser before getting to the heater. It may have slowed the heater getting warm slightly, but ensured that the vaporiser remained above -44 C.
96 Saab 900XS, AEB Leo, sold
93 Range Rover 4.2LSE, Lovato LovEco, sold
97 Range Rover 4.0SE, multipoint, sold
98 Ex-Police Range Rover 4.0, AEB Leo, daily motor
96 Range Rover 4.6HSE Ascot, AEB Leo, my spare


Proud member of the YCHJCYA2PDTHFH club.

LPGC
Installer
Posts: 4454
Joined: Wed Sep 29, 2010 6:01 pm
Location: Yorkshire
Contact:

Re: LPG or CNG: Why must we run Petrol at start-up?

#6 Post by LPGC » Thu Nov 25, 2021 12:13 am

The boiling point of LPG also depends on the pressure and we need it to boil (vaporise) - It doesn't boil in the tank even at 30C because when tank is at 30C the tank pressure is around 10bar.
There is a difference in the boiling point temperature between atmospheric pressure (mixer system reducer output pressure) and a sequential system output pressure which is usually between 0.9 and 2.1 bar above manifold pressure (and manifold pressure may reach atmospheric pressure... or higher on a turbo'd engine). At atmospheric pressure it's boiling point is -44C but at 1.3 bar above atmospheric it's boiling point is closer to -18C. Given that we might expect a 30C drop in temperature due to the cooling effect of boiling, a typical sequential setup running 1.3bar could see the reducer outputting liquid gas if the temp of gas in the tank is below 12C and/or the reducer hasn't been warmed by hot coolant from the engine (which itself might be at 12C when first started if ambient is 12C).

You need a good shot of antifreeze to prevent a reducer's water channels getting iced up (and hence blocked so no flow of heating water) if the reducer will ever see much colder than around -20C (or insert coolant freezing temp here). Even with a mixer system that might see LPG boil if the reducer is at -44C, if the reducer's heating channels are full of ice which blocks the flow of water it may see colder than that and still output liquid gas if the gas in the tank is cold. Of course if coolant only has enough concentration of antifreeze to prevent freezing down to -15C if the reducer ever does get below -15C the water channels will ice up and then the temp of the reducer will quickly fall. Series plumbing isn't always better than parallel plumbing... If it iced up with parallel but didn't ice up with series, then if coolant was changed (with higher concentration of antifreeze) it could be the higher concentration of antifreeze that made the difference between freezing and outputting liquid gas or not freezing and outputting vapour. Also, if the reducer is too high it may get less flow of coolant plumbed in parallel than if plumbed in series, but may get the same or even more flow if fitted lower and plumbed in parallel, series and parallel both have their pros and cons depending on the vehicle, flow ability of heater matrix's, etc. On some vehicles only one of the two is feasible.

With an electronic/pulsed LPG system the response of the injectors (opening and closing speed/times) can also change by quite a lot (in percentage terms) between cold and warm. I've previously mentioned injector response in terms of 'offset', could also term offset 'lag' or the difference between opening time and closing time. A hypothetical injector that takes 2ms to open and 2ms to close could be said to have zero offset unless the pulse duration is lower than 2ms (or the pulse length comes within 4ms of 'window'), the plunger in this injector might have enough inertia to reach fully open if pulsed for (say) 1.8ms to still open fully albeit reach fully open just a bit slower than if it were pulsed for the full 2ms... or then again it might not have enough inertia, and the 'then again' might also depend on the temperature. If this injector had a different opening speed and/or closing speed over a range of temperatures (which it will) it's offset will change over those temperatures. This hypothetical injector has 2ms opening and 2ms closing time... Now imagine we are expecting it to meter gas precisely when pulsed for just under 3ms (which may be the case on some installs).. If the opening time changes by 0.5 ms it has the potential to change the offset by 0.5ms, same for the closing time; If opening time doesn't change in the same direction and to the same extent over the temperature range as closing time there is potential for a 1ms total difference in response, which could equate to a difference in fuelling of more than 33% over a pulse duration just under 3ms.

Pressure also effects response and reducers tend to give slightly higher pressure output when they are cold (or when they are fed with vapour, or especially if they are fed with LPG vapour that contains air). The design of most injectors involves a plunger held closed by a spring and opened by an electric coil making an electro-magnet pulling iron in the plunger, the spring is relatively weak and it's force can be offset (effectively increased( by gas pressure acing to hold the plunger against it's seat during opening or relatively decreased by the flow of gas between plunger and seat slowing down the closing stage of the injector cycle. Lambda sensors tend to reflect a slightly leaner mixture on LPG than on petrol when cold or still warming up. Engines. Different spec engines have slightly different responses to running on LPG versus running on petrol, most will idle on very slightly less airflow when fully warmed up on LPG than on petrol

I won't need to say that cold gas is more dense than warm gas and that if everything else was unaffected by gas temp an injector would need to be pulsed for less duration with cold gas than with hot gas. But cold gas usually also implies cold reducer which normally implies cold engine. There is an interplay between temperature, pressure, injector response, lambda sensor response and engine response which tends to mean some interesting aspects occur (besides the more obvious choke effect and potential for the reducer to output liquid instead of vapour) when starting an engine with pulsed (particularly slave type) on gas when the engine is cold in cold weather.

It would be interesting to see development of an LPG ECU that did things the opposite way around to conventional systems... Instead of allowing complex modelling of fuelling with multiplier graphs or tables such setup might only allow a simple multiplier model... but allow a more complex model to describe injector response with a 3d table for injector opening response (temp/pressure/time) and another for closing response. A simple multiplier model would be all that was needed if injector response at varying temperature and pressure could be accurately described. It would still need adjusting after changing injectors but instead of changing the multiplier model the adjustment would be about describing the response of the new injectors.
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Post Reply