Featured Articles

Powerstroke High Pressure Oil Pump Woes
by Kim Lux
Copyright© 2000, 2001 Diesel Research Inc. - all rights reserved

Background

Ford Powerstroke engines, introduced in 1994, employ a unique fuel injection system, commonly termed HEUI, short for Hydraulic Electronic Unit Injection.

As the name implies, the HEUI injection system uses hydraulic energy to actuate the unit injectors. Not many people realize this, but a Powerstroke engine has TWO oil pumps: a low pressure lubrication pump, located near/in the front engine cover below the water pump and a high pressure hydraulic pump, located at the front of the engine vee, underneath the fuel filter assembly. If this is the first you've heard about the high pressure oil pump, don't feel bad: many times we've talked to Ford parts people that have no idea a Powerstroke has TWO oil pumps.

Pump Purpose

The purpose of the high pressure (HP) oil pump is to provide high pressure (up to 3,000 PSI) motor oil to the injectors. The pump is a seven piston axial (swash plate) design, of fixed displacement. Here are some links to swash plate pumps:

(Maintenance Resources Library)
The axial piston pump with rotating barrel by R. van den Brink
Axial Piston Pumps by Integrated Publishing

There are many more available on the internet.  The Powerstroke HP oil pump is fed low pressure oil from the lubricating pump via the HP oil reservoir, located at the front of the engine. Contrary to popular belief, the HP oil reservoir does not store high pressure oil: the oil in the reservoir is at lubrication pressure, 15 to 45(?) PSI , not injection oil pressure (500 to 2700ish PSI.) A more correct term for this reservoir would be HP oil pump reservoir, but everyone seems to leave out the "pump" part, thus leading some people to think the reservoir is holding 3,000 PSI oil.

The high pressure oil pump feeds two high pressure oil rails, one in each head, via two hoses. The output pressure of the pump is controlled by an electromagnetic pressure regulator mounted in the body of the pump, commonly termed the Injection Pressure Regulator (IPR), although it has other names. (Rail Pressure Control Valve, RPCV, if I remember correctly.)

There are at least 2 different high pressure oil pumps used on 99+ Powerstroke engines:

1999 engines built before December 12, 1998 use a pump common to the A250 T444e engines. 1999.5+ engines use a very similar pump, albeit with about 30% more displacement and a different swash plate/rear cover assembly.  There is some confusion as to the part numbers of these respective pumps, but it appears the 99.5+ pumps have a part number ending in "CRM" International appears to use a better part number system than Ford does.

A 1999 Powerstroke engine runs better with a 1999.5+ pump installed, although the parasitic power loss will be higher. A 1999.5 engine will run with a 1999 HP pump, but it will be somewhat performance limited. The same IPR assembly is used with both pumps.

Recommendation: if you ever get the HP oil pump replaced on a '99 PSD, attempt to get the '99.5+ pump. It will bolt right in.

The HP oil pump has a built in pressure relief valve factory set at 3800 to 4000 PSI to prevent damage to other HEUI components in the event of a short circuit to the IPR control circuit. (The IPR is controlled by a PWM low side driver.)

Pump Problems

During the course of our cackle kit development, we encountered a Powerstroke engine that idled roughly. After a significant troubleshooting effort, we determined the problem must lie within the high pressure oil pump. We developed several special tools that allowed us to measure and observe the regulator bypass flow of a pump while idling a Powerstroke engine. We determined the bypass flow on that particular truck to be both low in volume and very uneven: something was wrong with the HP oil pump.

We removed the pump from the engine and performed a disassembly/autopsy as illustrated in the following pictures.

Fig 1

Figure One showed the exploded parts view of a high pressure oil pump. From
right to left, we have the following parts:

a) pump drive gear and bolt
- the pump is NOT timed to the camshaft

b) below that, we have the pump drive shaft

c) to the left of that we have the brass cylinder, with the seven pistons (six in and one out)

d) directly behind the cylinder we have the pump body into which all these
pieces fit

e) to the left of the cylinder, we have the swash plate/rear pump housing

f) the bearing and rear seal

Fig 2

The HP pump on a PSD is an aluminum bodied swash plate pump. A swash pump is a cylinder with bores in it. Pistons are made to go in and out of the bores via a ramp. This ramp is called the swash plate. The pistons and the rotating cylinder assembly are brass and they are beautiful.

The drive gear connects to the pump shaft, which in turn turns the rotor via a spline. Why a spline? The rotor needs to be able to slide independently of the shaft to maintain a tight clearance.

The body is aluminum. Disassembling these pumps is very difficult the first time. The entire pump presses together. We had to machine several press tools, for both holding and pressing to successfully disassemble my pump. It was a very trying process.

Fig 3

The pistons have 2 mechanisms for ensuring that they retract fully:

a) small springs

b) a plate that holds a ball and socket connector that positively retracts each piston

The pistons compress automatically because the ball and socket mechanism follows the swash plate. As long as the pistons retract successfully, the pump will pump a full stroke of oil.

The plate that ensures the pistons move in and out of their bores, the piston "retraction plate", is held closely to the swash ramp via an internal c clip, like an internal snap ring, but without the eyelets to remove it.  Although there are springs on the pistons to help them retract, it appears as though they are not strong enough to do this alone, thus the retraction plate is used as well.

The internal c clip holds the retraction plate close to the swash ramp. (i.e.: the retraction plate sits between the swash ramp and the C clip.)

The groove that the c clip sits in is not very deep. The C clip on this pump had jumped out of the groove and became pushed up the bore on the fully retracted side. This allowed the retraction plate to float wherever the piston springs pushed the pistons. Wear patterns on the swash ramp show that the pistons were traveling less than 50% of their stroke as it was clear that the piston shoes were only in contact with a portion of the swash plate.

Cause Of the Low Volume/Uneven Bypass Flow

We traced the problem of low bypass volume and sporadic flow at idle to the displaced piston retainer plate ring near the swash plate of the pump.

A swash plate pump works like this:

• The shaft turns the rotor
• The turning of the rotor will cause the pistons to turn with it

The shoes on the pistons are held to the tilted swash plate via a piston retainer plate that is held close to the swash plate via the star shaped piston retainer plate

The piston retainer plate is held near the swash plate by a special snap ring, which sits in a groove in the bore of the swash plate.  Without the piston retainer plate ring, there was nothing to keep the pistons following the surface of the swash plate. During the pumping portion of the rotation, the swash plate will push the pistons towards the top of the slope, to the top of their stroke. Normally the piston retainer plate would pull the pistons back out of the bore as the plate followed the swash inclination. However, without the plate, there was nothing to pull the pistons back out of the bore. Engine lubrication oil pressure will act on the pistons to push them out of the bore, but under certain circumstances, this may not be sufficient, i.e.: hot idle, when oil pressure may fall to below 20 PSI. Furthermore, during engine starts, the lubricating oil pressure may take several cranks to build enough pressure to force the pistons to the bottom of their bore. There may be other operating conditions in which the pistons do not return to their bottom of stroke position.

Significance Of the HP Oil Pump to Proper HEUI Operation

If the famous HEUI injectors are the stars of the HEUI system, the oil pump is the supporting cast: without a steady, sufficient volume flow of oil to the injectors, the entire engine operation will be compromised.

The HEUI injection system is a pressure -time injection system, not a positive displacement system. The famous 12V Cummins B5.9 with its P7100 (in line) and VE (rotary) fuel pumps uses a positive displacement system: a cam in the high pressure fuel pump moves the (a) plunger in the (a) barrel a positive distance, thus displacing a fixed amount of fuel. Short of breaking the pump, it will move its volume of fuel to the injector tip.

The injectors in a HEUI system work on an entirely different principle: high pressure oil, at a set pressure (via the IPR) is allowed to act on the intensifier piston for a certain length of time via a solenoid valve on the top of the injector. If the pressure or time is increased, more fuel will be pumped. If the pressure is not the same from injector to injector, they will each pump different amounts of fuel, resulting in uneven engine operation.

(I'll talk about the significance, ramifications and effects of increasing fuel delivery by altering injector pulse widths and injection pressure via chips and "10K resistor tricks" in a future article.)

In summary , the ECM measures the HP oil pressure at one point in the rail and infrequently, assuming it is steady everywhere else and in between samples. In the meantime it issues commands to the IDM to fire the injectors assuming the oil pressure has not changed. With a low volume or erratically operating pump, the pressure will fluctuate more than it should and the engine will not operate in a smooth manner.

Furthermore, HP oil pumps with a displaced piston retainer plate ring are prone to operate erratically. Whenever the combination of variables (oil viscosity, engine (pump) speed, injector oil usage, etc) are such that the pump pistons are not delivering sufficient oil to the injectors, performance is going to be compromised. Due to the number of variables involved, this could occur at seemingly random intervals.

Disassembling/Assembling Powerstroke High Pressure Pumps

Lest one think that they may simply remove the pump from a Powerstroke and disassemble it to check if the piston retainer plate ring is displaced, you are in for a bit of a surprise: the Powerstroke pump is quite difficult to work on, requiring special tools for both disassembly and assembly. For example, it took 2 of us about 20 hours of tool fabrication and learning to disassemble and properly assemble our first Powerstroke high pressure pump.

There are no bolts holding the pump together: all the parts press together.  Furthermore, as far as we can tell, one must destroy the front seal of the pump in the process and we have yet to find a suitable replacement. Some people have told us that such pumps are not designed to be rebuilt, not without replacement of the significant components.

I've got one technical GEM for anyone removing the high pressure oil pump from a Powerstroke engine: REMOVE THE IPR BEFORE REMOVING THE PUMP FROM THE ENGINE. The reason for doing this is simple: if you drop the pump, there is a good chance the IPR will be bent and they are very difficult to straighten properly. Furthermore, they are expensive !

Troubleshooting High Pressure Pump Problems

It is quite difficult to properly diagnose Powerstroke HP pump problems without some special purpose tools. I'd like to give you a set of simple symptoms that point to a misbehaving high pressure oil pump, but even that is difficult to do because other components of the HEUI injection system can cause similar symptoms. Furthermore, we have reviewed the Ford troubleshooting procedures and observed the actions of Ford Diesel technicians. In our opinion they do not have the tools or procedures to identify a pump problem. (My apologies to Ford technicians...) For example, Ford Powerstroke technicians do not have any means of measuring pump flow, something that we think is critical to properly troubleshooting pump problems.

Here is a list of symptoms that would make us suspect a problem with the high pressure oil pump:

a) rough idle when hot, but generally OK when cold, although it may idle rough when cold as well

b) a stumble (not cackle) when the accelerator is pressed firmly in the 1000 to 1500 RPM range

c) the engine has difficulty keeping a steady idle speed WITH THE HVAC SYSTEM TURNED OFF

d) long crank time in cold temperatures (below freezing)

e) lack of acceleration on initial application of throttle when engine is first under load. (We are verifying this symptom.)

f) lack of acceleration in the first portion of accelerator movement. (We are verifying this symptom.)

Symptom C is probably most telling of a pump problem, but still not conclusive. However, if the pump output is sporadic, the pressure will be sporadic in the oil rail and the ECM will constantly be adjusting the oil pressure, resulting in an engine that never idles at a set speed.

These symptoms are not definitive: for example IPR, injector or injector seal problems can cause symptoms that appear similar to a HP oil pump problem.

The Ford procedure for troubleshooting the HP oil pump involves running the engine on only one cylinder bank and swapping the pump outputs. Such a test does not isolate the problem to the pump as a bad injector or injector seals in either bank can foul the results.

At Diesel Research, we use the following method(s) to isolate a HP oil pump problem:

a) using a specially built tool, remove the IPR from the back of the pump and replace it with a specially machined IPR plug. The IPR plug plugs the HP oil relief provided by the IPR. DO NOT RUN THE ENGINE THIS WAY.

B1) install a manual pressure regulator set at about 500 PSI in the HP oil supply circuit
-OR-
B2) using a special IPR relocation body, install the IPR outside of the pump body in the HP oil supply circuit.

In either case, route the regulator bypass flow into the oil fill tube on the passenger head.

c) start the engine. Remove the bypass hose from the fill tube and observe the oil flow. It should be relatively smooth. The bypass oil flow should be at least 750 ccs/ minute (99) / 1 liter per minute (99.5+). Anything less than this is cause for concern.

NOTE 1: some people will interject that one could have simply tested the pressure capabilities of the high pressure oil pump using some of the techniques outlined in the Injector O Ring article. Testing the pressure capability of the HP oil pump is not the same as testing its volume. Pumps may test fine pressure wise but not produce sufficient volume.

NOTE 2: The above test does not conclusively narrow the problem to the HP oil pump. Certain injector problems can cause injectors to use large amounts of oil and create what looks like a pump problem. The distinguishing factor is that such injectors will generally cause a bad hot start problem AFTER the first time the truck is put under load, but not necessarily before.  (Don't ask how I learned that... )

NOTE 3: Some people may say their Powerstroke engines have good full throttle power and thus they have a good pump. Lubricating oil pressure at full throttle on a PSD is around 50 PSI, probably enough to push the pump pistons to the bottom of their bores. Good full throttle power does not guarantee a good pump.

A better test for a faulty HP oil pump is as follows:

a) using a specially built pump mount, mount the pump in a hydraulic pump test stand. The PSD HP oil pump is not a standard SAE mount: the oil supply is fed to the pump via a port on the face of the pump. Several other ports are needed as well.

b) supply the pump with a low pressure oil supply, a few PSI at most, but not low enough to cavitate the pump. Measure the volume pumped at a set RPM (350 RPM or so.)

c) increase the supply pressure and repeat the measurement in b). 50 PSI.  Provided the pump was not cavitating originally, there should be no change in volume. If there is a change in volume pumped, it is likely that the pistons are not returning to the bottom of their stroke with the retainer plate.

There are several other tests as well, that should be performed on the HP oil pump if you've got it mounted on a test stand.

We've got another method for testing the HP pump, but it doesn't warrant explanation at this point.

Fixing the Displaced Piston Retainer Plate Ring

Our fix for the displaced piston retainer plate ring was to machine a sleeve that fit tightly in the bore of the swash plate cavity and yet did not interfere with the pistons. THIS IS NOT A TRIVIAL EXERCISE ! The sleeve sat above the piston retainer plate ring and held it firmly in its groove. The pump was run for a month and then disassembled and inspected. The piston retainer plate ring was found to be sitting nicely in its groove. The pump was tested for bypass volume in a running engine several times and found to be good.

We performed this fix because we wanted to prove that the low pump bypass volume at idle was caused by a problem within the pump itself. It is not economically feasible to perform this fix on a pump that is not operating properly. On the other hand, there is no guarantee that a new pump would perform any differently than a faulty one, because it too could be or could become faulty if the piston retainer plate ring slipped out of its groove.

Extent of the HP Oil Pump Problem

The pump the displaced piston retainer plate ring was found in showed no evidence of abnormal wear or damage. From wear marks, it appears the ring ran in the groove for a while and then popped out of the groove. My best *GUESS* is that the ring jumped out the groove during a period of pistons sticking in the bores, maybe during a cold engine start.

We have no idea how many PSD engines are suffering from this problem.  Furthermore, the problem is difficult to isolate without special tools. Yet, however, a PSD will not operate correctly with a faulty pump.

Aggravating Conditions For '99 Engines?

During an engine start up whereby the oil pressure in the HP oil reservoir is low, there is a quick fill check valve that will allow the reservoir to be quickly filled from the driver's side lifter gallery rather than from the (filtered) oil supply from the low pressure oil pump. This quick fill check valve can be found on the driver's side of the engine V, up near the HP oil reservoir. It is capped by a standard pipe plug. We suspect that the oil in the lifter gallery has the potential to be relatively unfiltered compared to the normal filtered supply, thus potentially allowing debris into the HP oil pump.

Interestingly, International (formerly Navistar) made a change to the gasket that seals the HP oil reservoir to the front engine cover between the 99 and 99.5+ models: the 99.5+ engine models have a gasket that employs a strainer to prevent larger debris from ever entering the HP oil pump. The gasket on the 99 reservoir does NOT employ the strainer. The gaskets are NOT interchangeable between the engines. The front engine cover is machined differently on the 99.5+ engine to facilitate seating the strainer gasket.

Conclusion

The HP oil pump on a PSD engine is critical to proper engine operation. It is not an easy task to troubleshoot a high pressure oil pump volume problem without proper tools. Diesel Research has found a problem in a high pressure oil pump that could exist in other engines. The HP oil pumps are difficult if not impossible for the average truck owner to disassemble. Diesel Research fixed the broken high pressure oil pump with a specially machined sleeve, but does not have an interest in testing or fixing HP oil pumps en masse.

Special Request

We are in need of several 99.5+ high pressure oil pumps. If anyone knows of a source for these pumps, new or used, we would be interested in hearing about it. We would also like to purchase a 99+ ECM.

Kim Lux
Diesel Research, Inc.