What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Boost Pump Operation

Boost pumps

These pumps all do the same thing.
Ours has the highest quality pump motor of the three options (Walbro).
Our pump has the lowest suction loss (by far) of the three options available because we use a better check valve.
Our recommendations may be more conservative because we like to make sure products are installed correctly and for the correct purposes.
This in no way represents a limitation of the product.
We have no failures of Boost Pumps in ten years of selling them.
I don't think any one else can make that claim.

Robert
 
Our pump has the lowest suction loss (by far) of the three options available because we use a better check valve.

As previously noted, that claim seems to be based on the old style AFP pump package. In fairness, perhaps it is time to update the claim (and the comparison on the EFII sales page) with measurements of the current pump package.

Past time really. This stored on my hard drive, a press release dating from 2013.

bj901y.jpg
 
Interesting discussion.

A few points that I haven’t seen discussed yet. Leaving EFII out and just referencing traditional carb/standard FI, operation, the engines themselves are massive pumps and will suck their own fuel, just not to the level needed for a high power setting. For high wing aircraft, gravity alone can suffice. I’m not an engineer, but I believe that there is also a positive tank pressure effect from the fuel tank vent ram air.

At takeoff power/take off pitch attitude in a low wing aircraft, a failure of the mechanical pump would likely cause an engine stoppage which is why boost pump use for take off is standard procedure. You would likely never know that the mechanical pump had failed until you turned the boost pump off, which is good to find out once at a safe altitude.

At a normal cruise power/attitude, a mechanical pump failure should not cause a complete stoppage. It did not the one time it happened to me in a carbed Piper.
It just caused a rough running lower power output with a low fuel flow indication and then normal operation was restored with the boost pump confirming the diagnosis.

At this point I would practice and teach that the aircraft is in failure mode, even with the boost pump operating normally, and that you need to be finding a suitable place to land where repairs can take place and not continuing another five hours just to get to your destination (unless you’re between Iceland and Scotland). A precautionary state, but less than an emergency.

The primary reason that boost pump usage is recommended for approach/landing is not for the arrival itself, but for the possibility of a high power go-around/missed approach.

In my 10 I do not use the boost pump for every tank switch, and I’ve never had a cough or stutter ever. I do, however have a top of climb flow when I make the first tank switch and I do check the functionality of both the fuel pump and the fuel pressure indicator. That also verifies that I didn’t forget to turn it off from takeoff, which, I don’t want to do on a regular basis, but don’t lose sleep over.
 
Last edited:
A Lyc engine won't suck fuel. Some amount of pressure is required for even a carb to work. Head pressure from a gravity fed system is more than enough but even with dihedral, a low wing RV probably won't ever have enough.
In the case of your Cherokee example, most like the pump didn't 100% fail and that is why it stayed running, though poorly.
 
Good information, thank you. It doesn’t really change things from a pilot operational standpoint, but is important to understand. By nature of their very design, the mechanical pumps are not something that can be tested/experimented with in flight. Both of you guys are much higher level subject matter experts on the technical side than I am and I appreciate and note the updated information.
 
Last edited:
Delving a little deeper:

Most hard failures of the mechanical pump on a Bendix type FI system will cause the engine to stop.

The quote above may well be true, but consider the following five failure modes, and the consequence of having a functional backup pump in each instance: (full disclosure: what follows is part of an upcoming series of posts I?m working on arguing for complete elimination of the engine-driven pump in some Bendix-type installations including my own.)

1- Actuator lever breaks- - Assuming it remains attached to the pump body, or falls harmlessly to the bottom of the sump, fuel flow through the injectors ceases immediately but is fully restored with activation of the backup pump.
2- Actuator return spring breaks- - Depending upon nature of the failure, flow to the injectors slows or stops, but is fully restored upon activation of the backup pump. In fact, pressure from the backup pump may actually step in as a partial substitute for the function of the failed spring.
3- Failure of one of the unidirectional valves that admit fuel into the diaphragm chamber or out to the fuel divider line- - Unless the valve failure has somehow physically blocked all fuel flow, full pressure to the injectors will be restored upon activation of the backup pump.
4- Rupture of the lower diaphragm- - Here is where things can get messy. Depending upon the nature of the rupture, flow through the injectors will be diminished or stop. Some of the fuel will pass into the chamber between the two diaphragms on the downstroke and some of that back into the lower chamber on the upstroke. Assuming the vent line intended to depressurize this area is properly plumbed to a good dumping point, some fuel may go overboard and continue doing so if the rupture is small and the engine keeps running. NOW- if you have decided to flip on the backup pump because the engine quit or went rough, or because you saw a pressure drop, any number of things can happen. Depending on the size of the rupture, you will now be dumping up to the full output of the electric pump overboard(or, ahem, into the engine compartment and onto the exhaust system if you never got around to that plumbing referenced above). There MAY be enough residual pressure due to flow restriction and/or a smallish rupture to get fuel to and through the injectors, but if power is not restored immediately you are in a potentially grim situation and it is time to shut off that pump and close the main fuel supply valve.
5- Rupture of the upper diaphragm- - The upper diaphragm is intended to keep engine oil out of the fuel section and fuel out of the crankcase. Its partial failure could go unnoticed for quite some time, but would generally result in some oil escaping through the vent between the diaphragms. A preflight checklist item worth adding is to verify that the dump tube is free of both oil and fuel residue. Simultaneous failure of both diaphragms is quite unlikely unless secondary to a major mechanical pump failure. However, if there were a previously undetected rupture of the upper when the lower fails, flipping on the backup pump would result in fuel being pumped into the crankcase as well as overboard and, possibly, through the injectors as well. Again, if power is not fully restored immediately, it is system shutdown time.

I?m not in a position even to guess about the relative statistical likelihood of any of the above failure modes but, happily, they are all proven to be extremely rare. Clearly, a reliable backup pump would help a lot in the first three, and have consequences ranging from helpful to disastrous in the other two.

All this is intended just as food for thought.

Personally, I?ll be installing a second light-speed ignition at the same time as the Airflow Performance FI on my O-360, along with a monster twin EarthX EXT900-VNT Battery system that will fully support an all-electric fuel system as well. At that point the engine driven pump goes away on my bird. I?ll be documenting the fully-redundant twin pump system with return lines that I?m now working on here within a few months. -Otis
 
My post was only responding to the other post where it was believed that an FI engine would run without fuel pressure somehow. The backup boost pump would supply that pressure.
 
Interesting. Lots of opinions, and not one little bit of temperature data.



Dan- I don?t think the concern is exceeding any particular temperature- that is only one condition contributing to fuel vaporization. In fact, outright failure of the pump itself is unlikely and not even much of a concern.

There is no means of extracting fuel vapor in systems without return lines, and these pumps cannot effectively pump foam or vapor. If the incoming fuel is already warm and the pump assembly is heating up due to continued use while recirculating nearly all of the fuel it is moving, you could reach a point where output pressure falls below the minimum level required to sustain operation of the metering valve, distribution block and injectors. As net flow through the FWF plumbing subsequently falls off, the situation grows even worse, and there is literally nothing you can do to remedy it. That is why you would want to operate at the highest reasonable fuel flows if you found yourself relying upon the boost pump after failure of the engine driven pump.

It might be possible to set up and test conditions required for this, but doing it safely would require very careful planning.

Engine driven pumps are actually very reliable, but their failure modes can be quite grim and, ironically, use of a boost pump during some of them will make the situation worse. I?m now assembling a fully redundant dual electric pump setup with no single-point failure modes for the Airflow Performance FI system I?ll be installing on my 7A next Winter. It will include fuel return lines and a duplex valve that switches both feed and return lines between tanks. I will be completely removing the engine driven pump, and will document all on VAF.

I will be watching and reading with interest. That sounds like an interesting set up. Thanks.
 
My post was only responding to the other post where it was believed that an FI engine would run without fuel pressure somehow. The backup boost pump would supply that pressure.

Yes, and I was confirming that your post was correct, but adding that engine driven pump failure modes can be complicated. You cannot simply say that the backup pump would supply pressure required to keep the engine running. Insuring that would require some very complicated valving that completely bypasses the engine driven pump.- Otis
 
Yes, and I was confirming that your post was correct, but adding that engine driven pump failure modes can be complicated. You cannot simply say that the backup pump would supply pressure required to keep the engine running. Insuring that would require some very complicated valving that completely bypasses the engine driven pump.- Otis

Agreed. Several different failure modes as you pointed out. Twin electric pumps would mostly eliminate that possibility as long as you had electrical power.
 
mechanical pump failure modes

Yes, and I was confirming that your post was correct, but adding that engine driven pump failure modes can be complicated. You cannot simply say that the backup pump would supply pressure required to keep the engine running. Insuring that would require some very complicated valving that completely bypasses the engine driven pump.- Otis
Hi Otis, do you have more info on these failure modes? I for one have an engine driven fuel pump that has a backup boost pump (andair) and while I can imagine some mechanical pump failures that would stop the fan from spinning, I've never heard of one happening. Thanks!
 
So I thought I would give an update since I started this thread.

In the end, I decided to return the Efii pump. Robert accepted the pump and refunded the purchase price. I thank him for that.

Yesterday I ordered a Airflow Performance pump form Don. I also ordered a set of lines and hoses for the pump from Tom at TS Flightlines. At the same time I ordered the install kit from Van's.

I'm very pleased with this new direction and am looking forward to installing the whole system.
 
Last edited:
Mike, I have this fuel pump in my 9A installation for the last 2 years of operation, in fact on their company web site has my mounting setup is the second pictures from the top. It runs very quiet so I added a bright green led when the switch is on. Never an issue many times I have even missed seeing the run light even after 5 minutes of running. Even with having the Boost Pump, If I lose the engine mechanical pump I'm looking for an airport to land.. Mark C. RV9A
 
Looks like this is now a dead thread, but I'll still add my two cents anyway: The issue of pump heat for a "boost pump" without a return path is real, but proper operation and understanding has shown that it is not significant. If it were, the bypass design of the EFii, Airflow, and Andair would be different. For example, instead of a return to the tank, you could add a small heat exchanger loop (maybe just a tube coil) to the either of the first two. Or you could put fins on the Andair body. But why add complication, weight, and complexity when not needed?

We know these systems WORK, so what follows is painfully unnecessary, but just to show some numbers in English units:

Assume nominal current draw of 4 amps at 12 volts = 48 watts
48 watts x 3.412 btu-hr/w = 164 btu/hr energy input. We will conservatively assume all the input energy goes into the gasoline. (This is fairly accurate at low temperature rises, but more conservative as the system temperature is higher since in reality some of the heat would escape to ambient via the pump body and connected tubing.) The specific heat of gasoline is about 0.51 btu/lb-F. Let's assume a fuel flow of only 1 gph. That's about 6 lbs /hr.

So, the gasoline temperature rise, conservatvely, is about:
164 btu/hr / (6 lb/hr x 0.51 btu/lb-F) = 54 degrees F steady state. Note that in this case, 54F is a fairly high rise, so there would be some heat loss to the surrounding air and it wouldn't really be this hot. But it's a worst case, as if you (stupidly) insulated the pump. If you are burning say 2 gph, it would only be a 27 degrees F rise, or at 10 gph, only 5.4 degrees F. But.... Don't leave power on and boost pump on without the engine running and think you can go to lunch! For that you would certainly need a cooler or return to the tank. Not sure what would fail, but something would.

Sorry for the long wind. Engineers are like that.

BTW, the Andair pump is cheaper this week if you order from England due the very favorable dollar to pound exchange rate. Maybe a buy opportunity until prices adjust.

Ted
 
Last edited:
FWIW... The electric pump in my Airflow Performance pump/filter assembly failed recently after about 650 hours on the meter. Normal takeoff/landing use (I fly a lot of short hops so it?s been on a lot) and occasional oops when I?ve forgotten to turn it off despite the annunciator light. It gave warning. I noticed the pump seemed noisy, and a couple of passengers commented on it. Then the fuse 10 amps) blew a couple of times. Then once in the ground it didn?t run when I turned on the switch. The light came on but no noise. Quickly cycled the switch and it did come on.

That was the last straw. Grounded, ordered new pump from AFP. The new one is so quiet I can?t hear it when the engine is running. Based on my automotive experience the failure seemed typical- they usually don?t fail while they are running. Instead they just don?t turn on and/or they blow the circuit protection. I had expected the pump to last longer than it did, but stuff happens. I don?t have a good sense of mean TBF for high pressure EFI-type pumps in aviation use. In any case I welcome the (slight) reduction in cockpit noise...
 
I had 5000 hours on a Walbro EFI type pump in our shop car. Still quiet and working fine when the car was sold in 2016.
 
I currently have a dukes but intend on replacing it with an AFP when this unit dies. When I initially turn on the pump for a cold start, or a hot start sitting on a hot ramp you can hear it cavitate at first as it gets its prime. I expect this will eventually kill it. I keep this in mind and always listen for a change in noise or pressure loss to signal when it?s time. I think it?s just a product of the design.
 
I had 5000 hours on a Walbro EFI type pump in our shop car. Still quiet and working fine when the car was sold in 2016.

120k miles over 17 years on a Ford-spec Bosch frame mounted pump on my EFI-converted ancient Ford Bronco. Still good. I don?t know who makes the pump AFP sells. No markings other than a part number. All the Walbro and Bosch pumps I?ve seen had the manufacturer?s name on them.
 
120k miles over 17 years on a Ford-spec Bosch frame mounted pump on my EFI-converted ancient Ford Bronco. Still good. I don?t know who makes the pump AFP sells. No markings other than a part number. All the Walbro and Bosch pumps I?ve seen had the manufacturer?s name on them.

This it?
https://www.amazon.com/Airtex-E8248-Performance-Multi-Port-Applications/dp/B000DN7VME/ref=sr_1_8?ie=UTF8&qid=1533920211&sr=8-8&keywords=airtex+fuel+pump

Anyone peel off the AFP sticker & look under it?
 
Or maybe Airtex E2315 - a roller vane automotive pump. That pump capacity is 38 gph at 0 psi, max discharge about 120 psi, obviously controlled by the relief path. Too bad we don't have side by side tests to compare reliability of all the different pumps in aircraft service.

One thing I do know is the obvious - any pump can be damaged by adverse conditions. I have never had an auto pump fail in automotive service except once (and I typically drive my cars to about 200,000 miles plus). The one that failed was run out of gas at about 50k miles. That was an older throttle body injection car. The pump ran, but no longer could overcome vapor lock and it wouldn't start if hot (return line had a back pressure control that wouldn't open with reduced pressure). Given that the pump can be damaged by any number of things, we just have to understand them, operate correctly as best we can, and observe any degradation. Fortunately, I think damage (except maybe dirt) will mostly affect them on the ground unless you run a tank dry with a pump can't tolerate it. (that point might be worth exploring -- not sure if any publish dry run time information.)

Ted
 
Last edited:
I think AFP uses or used Airtex pumps (someone should verify that) which have a limited lifetime warranty from the manufacturer. Interestingly, they only warranty their mechanical pumps for 12 months. Vehicles are the word they use, an airplane is a vehicle technically right? ;)

An alternative engine vendor supplied Airtex pumps with their packages and I've heard of a fair number of failures of those where they start to growl and make metal.

I've owned a bunch of pretty high mileage EFI cars and worked on many others in my repair business. Saw one Nissan pump fail at 325,000 km, one BMW at 360,000 km so those are at something over 200,000 miles. Probably over 5000 hours in both cases with usual average speeds and idling.

I've seen a bunch of ancient (early '70s) EFI D Jet Volvos in the bone yards with some really crusty pumps (externally mounted under the car). From the condition of them, they had not been changed for a very long time if ever. Two cars had 290,000 to 430,000 MILES on the odos. If the latter one was original, that's certainly over 10,000 hours. I took both pumps and used them on a test bench for years more. Still worked fine.

Bosch had some massive pre-filters on these cars which may be the secret to longevity since lots of folks never change filters.

I've sold hundreds of the Walbro pumps over about 15 years and nobody has ever reported a failure unless they were mounted incorrectly (those two failed in a few hours sucking a lot of air). Air or junk going through them will severely affect lifespan of roller vane pumps.
 
I think AFP uses or used Airtex pumps (someone should verify that) which have a limited lifetime warranty from the manufacturer. Interestingly, they only warranty their mechanical pumps for 12 months. Vehicles are the word they use, an airplane is a vehicle technically right? ;)

An alternative engine vendor supplied Airtex pumps with their packages and I've heard of a fair number of failures of those where they start to growl and make metal.

I've owned a bunch of pretty high mileage EFI cars and worked on many others in my repair business. Saw one Nissan pump fail at 325,000 km, one BMW at 360,000 km so those are at something over 200,000 miles. Probably over 5000 hours in both cases with usual average speeds and idling.

I've seen a bunch of ancient (early '70s) EFI D Jet Volvos in the bone yards with some really crusty pumps (externally mounted under the car). From the condition of them, they had not been changed for a very long time if ever. Two cars had 290,000 to 430,000 MILES on the odos. If the latter one was original, that's certainly over 10,000 hours. I took both pumps and used them on a test bench for years more. Still worked fine.

Bosch had some massive pre-filters on these cars which may be the secret to longevity since lots of folks never change filters.

I've sold hundreds of the Walbro pumps over about 15 years and nobody has ever reported a failure unless they were mounted incorrectly (those two failed in a few hours sucking a lot of air). Air or junk going through them will severely affect lifespan of roller vane pumps.

I bought a used AFP setup (old style). The pump had failed and i had to replace it. It was an airtex pump.

Larry
 
So you have an engine driven pump (Mechanical) and an electrical driven pump as a secondary boost pump. The only time the boost pump should be ON is during engine prime/starting and an emergency should the primary (Mechanical) pump fail. Keep in mind, anytime you compress a liquid it will get warm. However, your fuel injected system should be under 50 psi. Just add an indicator panel light to the secondary pump switch so you have a visual indication that the pump is ON. That should prevent you from accidently leaving it on for extended periods of time.
 
So you have an engine driven pump (Mechanical) and an electrical driven pump as a secondary boost pump. The only time the boost pump should be ON is during engine prime/starting and an emergency should the primary (Mechanical) pump fail. Keep in mind, anytime you compress a liquid it will get warm. However, your fuel injected system should be under 50 psi. Just add an indicator panel light to the secondary pump switch so you have a visual indication that the pump is ON. That should prevent you from accidently leaving it on for extended periods of time.

Lycoming begs to differ with you:

The conditions under which Lycoming recommends operation of the fuel boost pump are as follows:
1. Every takeoff.
2. Climb after takeoff unless Pilot?s Operating Handbook says it is not necessary.
3. When switching fuel selectors from one separate fuel tank to another, the fuel boost pump should be ?on? in the new tank until the operator is assured there will be no interruption of the fuel flow.
4. Every landing approach.
5. Any time the fuel pressure is fluctuating, and the engine is affected by the fluctuation.
6. Hot weather, hot engine ground operation where fuel vapor problems cause erratic engine operation.
7. Some General Aviation aircraft require the use of the fuel boost pump during high-altitude flight. This will be spelled out in the Pilot?s Operating Handbook.
8. If the engine-mounted fuel pump fails.
If the fuel boost pump is used during ground operation, don?t fail to check the condition of the engine-mounted fuel pump before takeoff by turning the boost pump off briefly, and then back ?on? for takeoff. If the engine-mounted pump has failed, it would be safer to know that on the ground rather than in the air when the fuel boost pump is turned ?off.?
When in doubt, do the safest thing and use the fuel boost pump with Lycoming engines. Don?t be ?stingy? with the boost pump. In most cases, they last the overhaul life of the engine, and are then exchanged or overhauled themselves. AS A REMINDER, the airframe Pilot?s Operating Handbook is the authority if boost pump information is spelled out in it.
 
You can even add, if leaving the planet turn on the boost pump. I understand what Lycoming says, I use it as needed. The point was, its only used intermittently, not for the entire flight so don't worry about the pump being turned on. However, RV7A Flyer I understand your attention to detail and by the BOOK response. After 35 years of flying military and civilian aircraft and with over 10 years as a Military test pilot I stand correct to your post, Use the boost pump for everything LYCOMING has ever recommended. Thanks for the feedback RV7A Flyer.
 
Back
Top