At 170K it can just be rodent fatigue under the hood. Likely you can verify the rodent fatigue by having an engine compression test done on all four or six cylinders of your engine. If you're lucky the low compression is due to leaky valves and a valve job can restore full power. Since all Subaru use a boxster design engine though there is a possibility that your cylinder sleeves have uneven wear due to gravity. The old air cooled VW bug was prone to this problem, too. It was relatively easy to replace the individual cylinder jugs of a bug with after market or junk parts to get them back to life. I have no idea how easy or complicated a Subaru will be.

Have checked compression and it is at acceptable range.

This does not happen at the beginning of the mountain which is steeper, it is after it has been pulling for a while.

I thought there maybe a sensor that is leaning the motor out and causeing the reduction of power.

Also have had both timing belts replaced about 15K ago.

Have no problem when going up steep hills on the highway, just after extended period over the mountain to work which is about 5 miles up the mountain, and it starts around 3 miles up.

This is the 4cyl boxer engine.

What I was asking for was the elevation above sea level it starts at and does it do it going both ways as in gains power once it gets below that elevation?

How high of elevation are you at when it starts loosing power and after how many miles does it happen?

Is it running hot when it looses power also?

A lot of vehicles are tuned only to run at normal elevations to meet emission's standards and once you get high enough in elevation that drop in air pressure and density is enough to make the computer turn down the power due to it having reached a A/F ratio control program limit.

The old carburetted engines gradually lost power as elevation increased and would start to blow black smoke out the exhaust once they got high enough due to the A/F ratio getting too far off whereas fuel injected engines just keep adjusting the A/F ratio until they reach a program limit of one sort or another at which they then can suddenly change performance depending on what the programming is set to do.

Without knowing the life long maintenance of the engine I personally would not consider 170K miles worn out yet or at least on Fords thats about half way to worn out! I have ran several now to 250k+ and the engines were still solid and did not use oil but the vehicles were falling apart around them!

The engine does not heat up at all, it stays at it's normal operating temp.

I was told the OX sensor and also the vacumn sensor could do this, then I was told these could not cause the problem.

To good of gas mileage to think of getting rid of it!!!

Also there are times this condition does not happen, it is rare but sometimes there is no loss of power???

the computer should compensate for the difference in altitude. try disconnecting the battery ground wire then reconnect it. that should reset the computer. if you still have same problem, take it to the shop.

You haven't said how much the altitude changes, but any engine produces less power as air density is reduced (higher altitude), although turbocharged engines are affected less. Have you changed the air and fuel filters and checked the fuel pump?

I am thinking Oxygen Sensor.

When was that last replaced?

You should be able to have that tested and it is not expensive to replace.

Well you can figure a 4% loss of HP per 1000' of altitude....add to that perhaps cooler temperatures causing slight deflation of tires, and any unknown factors could be critical at max output.....

I had a similar issue with a '86 Chevy Nova (or was it a Prizm). Anyway, it turned out to be a fouled catalytic converter. If possible, (temporarily) disconnect the exhaust (or tail) pipe between the engine and the converter and see if that makes a difference. However, be aware that potentially deadly CO fumes will enter the cabin if you drive far and/or idle at stop - so keep all the windows open and reconnect the converter (or replace, if indicated) at your earliest opportunity.

Seems like the most obvious place to start, no point in looking for an obscure solution at this stage. GA from me.

GA

Or fuel pump, not to forget that some systems have two pumps. One in the tank. When one fails, the other is simply not powerful enough to supply fuel when the nose is higher than the tail.

A worn out pump will give the same symptoms....

A blocked filter can also be part of the problem, that is, not checking both pumps but ONLY replacing the filter may fix it for a time, or improve things....

So the OP should get the filter replaced and check both/single pump(s).....

the air filter might need replace too. the less dense the air, the more air need to provide provide enough oxygen.

Often when a fuel pump gets old it can't put out enough fuel and when going up a hill; the demand for fuel is greater than it can produce so the power disappears. This also could show up at high speeds when more fuel is needed. It is a good idea to replace air and fuel filters also.

Malcomb is right...Last three cars i owned are subaru (impreza)...I had similar issue on one...change your fuel filter...if not, check/drain the the fuel tank sump. It could be that dirt build up is sitting on the bottom and when going on a long uphill drive the crap sitting on the bottom is swirled around and jamming fuel intake.

Just on a side note...what fuel are you using? E10? premium? subaru recommends 95 octane or higher. Over time, you may be compromising efficiency.

Perhaps a pressure sensor is gummed up, there are a number of air pressure related functions .... MAP sensor, fuel pressure to the injectors is referenced to the manifold pressure, and probably more the newer the car......

Do you had an in-tank fuel pump? If so check that it is pumping ok - many fuel injected cars have a low pressure high volume pump in or near the tank followed by a high pressure, low volume pump feeding the injectors. failure of the in-tank pump can give some strange symptoms.

I've read through the replies and can't see if you have said whether it's petrol or diesel, If it's petrol then a combination of air filter and fuel mixture would be my guess,get it put on an analyzer to check for air/fuel mixture, it will also show up any other faults so it's worth the expense.

Bazzer

I had a similar condition on a Ford Fiesta and it was the exhaust plugged from the converter coming apart and bits the size of our thumb jamming the pipe. Easy test is to loosen a flange if there is one and make the same drive. Noisy but if it changes symptoms you may have it. Another method is to drill small hole and install nipple so that you can install a pressure gauge and monitor exhaust back pressure. I understand 2-3 psi is normal above this you can expect losses in efficiency. Some vehicles have double wall pipes and the inner can collapse due to gasses between layers expanding and forcing inner pipe to collapse.

Fuel pump. As some has commented the demand for fuel is greater. Also the head pressure on the line is increased. On a incline the engine is much higher.

I would check the age/condition of the fuel and air filters. If they are acceptable, connect a vacuum gauge to the engine on the next hill climb. If it is the exhaust restricting the engine as the additional power is needed, the manifold vacuum will will turn to manifold pressure-from residual exhaust pressure. To test, simply create an exhaust leak before the cat, or wherever you think the exhaust may be restricting. The exhaust leak can be as simple as removing an O2 sensor, or loosening up the flange bolts a little. Good luck.

One more thing to consider - that the timing will retard if the knock/ping sensors are going off - and a lessening degree of spark timing advance is usually a lessening of power - without 'missing'. Does thinner air tend to increase the pinging? (I don't know.) Perhaps this is the effect that a failing oxygen sensor has (falsely reading less oxygen than 'real' => too lean?). Perhaps Daffy is on-mark to question the octane.

I don't see how this can be tested in the shop unless the computer stores 'max-min's, they are cleared before a trip up the hill and looked at right away afterward. You might want to see if you can borrow a scan tool to show the timing as you drive.

A scan gauge is invaluable for checking out this sort of thing. You can see throttle position, timing, fuel flow, etc, etc as well as whether sensors are working correctly.

By pure guess, I'd think that the engine is tending to overheat slightly, leading to a little pinging, which will cause the timing to be automatically retarded. This can cause additional overheating, and can cause additional pinging tendency, which will cause the timing to be retarded even further.

The root cause could be carbon build up. The carbon causes an increase in compression ratio and an insulting effect, both of which can lead to pinging -- which in the old days you would hear, but now the timing is automatically compensated at the onset of pinging. Try high octane fuel, and see if there is any difference.

Scanning is most direct way to know what's going on.

Other possibilties: Torque converter lockup failing to unlock? If it is a manual trans car, clutch slip would be a possibility -- but hopefully you wold recognize this pretty instantly: the engine speeds up but the car does not.

I am assuming this is a new problem. Any car losses power with altitude.

The Suburu knock sensors are very sensitive, picking up not just the onset of pinging but also loud voices. So if you shout insults at your engine, such as "Even a Yugo could produce more power than this!!!" the ECU will retard the timing, making the car go even slower. Some have accused Suburus of being passive aggressive because of this trait.

LynnDoor enterprises sells a set of little ear muffs that can be put over the knock sensor, shielding it from shouts but still allowing it to pick up engine shock waves.

Good one.

I assume that going up hill, you depress the accelerator to minimize slowing down ?

A weak catalytic converter could cause what you're describing.

My thoughts exactly at post 2 and my follow up question about the elevation in relation to sea level in post 7.

I have had two similar problems, both caused by fuel line restrictions. One was caused by a ding in the fuel line and the other by a loose particle. In both cases the car would run normally on the straight and level, but when enough gas was being used that the float level became so low that the gas would not feed, the power dropped off.

And there could be a loose bowling ball in the trunk. When driving up the steep hill it rolls over the hole in the floor of the trunk, that is also the roof of the gas tank. That causes the tank to loose the vent, and after time the fuel pump can not overcome the additional work required to bring fuel from the vacuum to the injectors.

Or, there could be a large rubber ball in the wheel well. And when he drives up the mountain, the ball falls against the tire and rubs severely. The extra drag appears as low power.

We can "what if" this to death, but at some point he must start buying random parts, or diagnose.

Maybe its got "Inverted Momentum" issues.

Is that where all of the Higg's Bosons are hiding?

Could be!

I wonder how their equations would come out if they used the wrong time and point references for the energy states of each particle and related sub particle after the collision?

I would bet there would be lots of FTL events that way!

I suggest you check the engine vacuum and verify if it is at the proper level/value for your altitude.

With a minimum amount of engenuity and a portable vacuum gauge, you can also adapt and monitor the vacuum of the engine during operation by extending the connector hose and bringing the gauge into the cabin where it is visible at all times.

If the vacuum is low:

1. You most likely have a vacuum operated component such as the air conditioning accumulator/reservoir, transmission vacuum modulator, or one of the many vacuum dependent emissions devices such as the EGR valve that is not sealing properly. (The last is very common with older vehicles with high mileage due to sludge buildup and wear of the valve seat and/or valve disc or electric solenoid coil failure so I would start there.)

One of the telltale signs of this failure in vehicles with vacuum operated AC controls is that the AC flow diverter gate will cycle and move the outlet air from the floor-to-dash or dash-to-windshield ports whenever the engine is put under load such as going up a grade or pulling a trailer without the control selector(s) being repositioned.

Simply attach the vacuum gauge to the engine then use a pair of hose crimping pliers to crimp the vacuum supply hose of each device until there is a noticable change in the vacuum level then remove the device and clean or replace as necessary. (Vise grip liers work well but be careful not to overcrimp and danage the hose(s).)

You may have more than one failed device so be sure to continue checking all vacuum operated devices after repair/replacement of the first and so on.

or

2. The engine is worn out and the compression is too low. (I Hope not.)

Good luck!

Turn off your AC while riding up to mountain...

Try changing down a gear.

Only noticeable horsepower robbing experience I have ever had due to altitude was on a 100cc Honda scooter climbing Doi Inthanan in Thailand.

Very very noticeable. Nothing that a bit of right wrist and a lower gear didn't overcome.

I figure there are some variables that OP hasn't fully transmitted to us.

I also would check the engine´s AIR intake: are the filter / filter canister clean?

Start looking for the trouble where it does not cost money: Loosen or remove the feul tank cap. When air can not enter the tank freely this trouble as you explain, can not be found in the workshop because it only happens when you have driven a while or when more feul is used as when you are driving uphil.

If that is the trouble, that is a really old car as for the last (at least) 30 years or so, have a special way of catching the petrol fumes in charcoal and reusing them in the engine......therefore no vent hole in the cap.....

But maybe some manufacturers were slower in adding this feature than others.....so you may be still right, though I cannot see why it doesn't also happen on the flat as well.......that would be a strange dividing line......

To my mind, it most probably has to be something that affects the fuel supply when driving at an angle upwards, as this for example, makes the fuel pump(s) work harder, especially if a fuel filter is really almost blocked. Its also a fault that I had to fix on a friend's Mother's car. Probably about 1978....I just remembered, that was a Toyota!!! As if that makes a difference...

At least that's my guess.....

Andy, You are right....but this is what just happened to my 1998 1300 Daihatzu Terios with 270 000km on the clock. Also changed the blocked front silencer with home-made one.

Only found this trouble after I replaced the feul filter, petrol pump and three relays!

As soon as I find the time I have to find the bockage of a pipe though which the feul tank is supplied with air.

Does any-one know where I can find a good wiring diagram for the Terios?

I have found the problem with the 2004 Subaru!!!

Knock Sensor was bad: Mechanic said it was retaring the timing long before it sould have.

Replaced and now it runs like it use to.

Thaks all of the ideas and the checks that lead up to the solution.

Sweet fix.

Learning all the time, remembering is the challenge...

I wonder why the buggered knock sensor wasn't detected by the OBD and issue a trouble code?

Do you know how the mechanic came to his/her (very accurate) diagnosis?

Thank you for letting us know - We have all learned a little