Plantoil/diesel conversion basics
Sign up Latest Topics Chat
 
 
 


Reply
  Author   Comment  
danalinscott

Moderator
Registered:
Posts: 934
Reply with quote  #1 
(Under construction)

Since no clear definition has been provided before this I will give it a try.

"Purge time" is the amount of time at idle that it takes a two tank conversion to purge VO fuel (replacing it with (bio)diesel fuel) in the fuel system from where the two fuels "merge" prior to the IP to the return fuel valve.
 
Questions?
Comments?

__________________
Dana danalinscott@yahoo.com
JohnO

Registered:
Posts: 16
Reply with quote  #2 

I agree that purge time at idle is good to know, but most users might prefer to know how much time/how far to drive in order to be purged before they park and shut down. Flow rate at idle may be so low that purging at idle is impractical (as I discovered with my Mazda, which would not adequately purge at idle in 20 minutes).
In which case the definition would reflect some normal engine load or rpm. It might also be limited to the fuel system from IP inlet to fuel return connection, accounting for the external parts (valve, hose, heat exchangers, filters, etc) separately. I can also easily imagine a graph showing no-load purge time vs rpm, possibly with additional lines for the external parts.

An additional thought - I suspect that each IP may have some unique amount of fuel that must flow through it to thoroughly purge, in which case a definition might benefit from being based on a volume and flow rate, rather than simply time. Due to mixing and dilution, this amount will be greater than the internal volume. For example, a 4-cylinder VW IP may hold 500cc (I'm guessing) but might need to flow 1500cc before it is "purged".

Purging is normally done with the return going back to a tank, rather than looped. This may be based on a bad assumption. I'll leave it at that for now.



__________________
JohnO
JohnO

Registered:
Posts: 16
Reply with quote  #3 

I've noticed a huge difference between my VW and Mazda's purge times/distances. They're both home-made 2-tank systems, very similar to each other in most details (since I designed and built both systems). This winter I've been running B100 in the VW heated tank, #2 diesel in the main tank. This allows easy startups on cold (20F) mornings. At that temperature the B100 is nearly solid and will not start. This is true in either vehicle. I used the same method in the Mazda for the previous 2 winters.
When returning home from work, the VW can be switched to the main tank (beginning the purge) when I pull up to my gate, which is less than 100 feet from where it gets parked. Less than 2 minutes later I shut it off. It starts easily the next morning, smelling of normal diesel. I initially was switching to the main tank about 3 miles from home, as I had done when driving the Mazda. I had learned by hard experience that the Mazda would become hard to start the next morning if I waited until 1 mile from home, and would not start at all at 20F if I had switched at the gate the night before.
My conclusion is the VW purges much much faster than the Mazda. I also suspect that it purges quite quickly even at idle, whereas the Mazda once idled for 20 minutes after switching back to diesel, yet did not start the next morning.

The "real-world" requirement for purging is the ability to easily start when cold. How do we define that in a useful way? The problem would be easier if all engines are able to start equally well with a predetermined residual percentage of svo. Is this true? If so, then finding an allowable percentage will allow direct measurement of svo dilution.

Do people understand that the purging process doesn't push the svo out of the system in a slug, but rather it dilutes it and the percentage gradually declines? This would imply a time constant might be a better way to understand the purge "time", which would be the time needed to reduce the svo content to some percentage (1/3 v/v?).

I'll use my VW as an example, making up some numbers for illustration:
Time constant at idle for svo in return line to decline to 33% by volume = 60 seconds
Allowable cold start svo % = 10%
Time needed to reach 10% svo concentration = 2 time constants (33% x 33%) = 2 minutes.

Time constant during driving is assumed to be half that of idling, so 1 minute driving will equal 2 minutes idling. This only applies to my VW, not the Mazda. Driving time constants are soft. Load and speed each have strong effects on return line flow rates, and combustion demand.

JohnO


__________________
JohnO
danalinscott

Moderator
Registered:
Posts: 934
Reply with quote  #4 

Quote:
The "real-world" requirement for purging is the ability to easily start when cold. How do we define that in a useful way?


Thanks for commenting on this.
The "definition" I provided was intended merely as a starting point.

Possibly the best definition requires an agreed upon definition of several things. For example.

When is an engine "purged".
You have suggested that this might be
Quote:
the ability to easily start when cold
and I think that definition has merit. But..what is "cold" must then be defined.

For the last few years I have  thought of "fully or completely purged" to mean that the fuel system from the diesel filter to the injector return line is filled with  diesel fuel uncontaminated with VO.  IMO all other conditions represent a "partially purged" condition.

I believe that Chris Goodwin (Frybrid) pioneered the use of fuel dye to determine how long it takes a particular engine/conversion to reach a "purged" state. And this may be the "standard practice" for doing so currently.  The procedure as I understand it is to use a clear return line to determine when the dyed fuel has replaced the undyed fuel in the injector lines. (or vice verse).  This however is very difficult to accomplish safely while the vehicle is begin used normally (e.g. being driven down the road at normal speeds).

Perhaps the next step is to determine if there is a way this procedure can be modified to allow that?  If THAT can be accomplished it will be much simpler to establish a practical, firm, and clear universal definition for "purged".

Has anyone yet done that?

 

__________________
Dana danalinscott@yahoo.com
JohnO

Registered:
Posts: 16
Reply with quote  #5 
You may be familiar with the concept of a "time constant"? Purge could be defined in terms of a similar "purge constant". It would be the volume, or time, or miles necessary for 2/3 of the fuel in the injection system to be displaced. Then it's a matter of testing to determine how many time constants are needed to obtain reliable cold starts without engine deposits.

This brings to mind assembling a matrix of common component volumes, hose volumes per length, IP internal volumes and their pumping rates. Then you would add up the internal volumes involved in a purge, and the pumping rate, and possibly the fuel characteristics (some might require more "constant" for a reliable cold start than others), and apply the "purge constant".

Here's the theory I'm assuming: the arriving new fuel doesn't displace 100% of the departing old fuel as it enters the IP. Instead, it dilutes the old fuel, mixing with it in the mechanism. This means the % of old fuel, the ratio of old to new, declines with time, rather than being displaced abruptly by the end of some time. Carried to an extreme, we could say that no matter how lon gyou purge, there will always remain some traice amount of the old fuel. We just want that remaining quantity to be so low that it doesn't cause any problem.

I will try to find time to experiment on my Jetta to see if I can determine it's "purge constant".
Cheers,
JohnO

__________________
JohnO
danalinscott

Moderator
Registered:
Posts: 934
Reply with quote  #6 
Quote:
Here's the theory I'm assuming: the arriving new fuel doesn't displace 100% of the departing old fuel as it enters the IP. Instead, it dilutes the old fuel, mixing with it in the mechanism.


I don't think it is safe to make that assumption.
My reasoning is this:
The volume of fuel in most IPs is fairly small so the "dilution" factor is small. More like purging a section of fuel line than a filter.

For the purpose of discussion:
I think the most significant factors in determining purge time is the volume of "spill fuel" that the injectors produce in a given period of time  and how much volume of fuel is contained in the "common area" of the conversions fuel system (where VO is replaced by diesel fuel during a purge). This "common area" is usually composed of the section of the fuel line from the point the two fuels seperate feed lines meet to the point after the injectors where they diverge or if no such second point exists to the spill line of of the last injector in the to become purged. 

For clarity:
Perhaps we could use a term like "spill rate" (cc's per second at a set no load RPM) to denote the first and "purge volume" (total cc's contained in the "common area" of the fuel system to denote the second.

It might be possible to determine purge time (at a given no load rpm) by simply dividing the purge volume by the spill rate.

What do you think?

__________________
Dana danalinscott@yahoo.com
JohnO

Registered:
Posts: 16
Reply with quote  #7 
My VW has a remarkably low "spill rate", to judge by the flow visible in the transparent spill lines between injectors and to the Tee. A relatively large volume enters the Tee from the IP return port. This is unusual in my experience with diesel farm equipment (and my Rover), where it is more common to see the majority of returning fuel coming from the injector spill pipes, and not much at all from the IP.

At first glance, the differences would not make any obvious difference to purge times - it's just a matter of waiting for the combined IP and Injector returns to flush out the Veg oil. Simple, right?

No, it's not quite that simple, especially with respect to the VW. The IP appears to purge very quickly, amazingly quickly. Unfortunately this may give a false sense that the purge is complete before the fuel in the injector lines has been used up. Remember that these engines barely even sip fuel at idle, so it is possible for the injectors to still be filled with unpurged SVO, so the next cold startup would begin with the injectors trying to spray cold svo.

The opposite situation could exist in other diesel engines: the injectors might purge quickly, but the IP could still be flowing svo at shutdown. In that case, the exhaust would smell like the engine had purged, but the next cold startup could be with cold svo in the IP, trying to pump it to the injectors.

The "simple" solution to this range of purge rates is to measure the IP return flow separate from the injector return flow. You still need to know how much needs to flow from each circuit to completely purge the system.

This needs a bit more thought and investigation to come up with precise answers, but in the mean time it would be nice to come up with some general rules for determing minimum safe purge methods. A subjective one that comes to mind is to conduct a poll, and assemble a data base organized by vehicle/engine/year, etc. I would give more weight to operators who have accumulated high mileages, using a particular purge technique, rather than operators who successfully use shorter purges but have less mileage.

My 87 Jetta 1.6 IDI turbo diesel accumulated roughly 200,000 miles between two owners using a 2 to 3 mile purge. The engine was retired at 300,000 miles, but the original IP continues on in the "new" engine. I now use a 1-mile purge, with apparent success.

__________________
JohnO
Previous Topic | Next Topic
Print
Reply

Quick Navigation:


Create your own forum with Website Toolbox!

This forum is brought to you by A proud sponsor of Invest in the future.