Reciprocating Gas Engine Compressors

1Q. What is a good indicator that the engine foundation is adequate?

1A. The movement of the engine on the foundation should be small, usually less than 0.005 inch total movement at any bolt in all three axis while under operation. Bearing problems or excessive crankweb deflection measurements are indications of possible alignment problems.

2Q. Why do the bolts keep breaking on the distance piece to engine frame on the compressor cylinders.

2A. Usually it is because the bolts loosen because they are short and have very little "stretch". Once they loosen they see the full cyclic fatigue load and fail. Frames are usually cast iron and the maintenance crew doesn't want to tighten up too much and pull the threads out of the low strength cast iron. One approach used is to tighten the bolts to 30,000 psi nominal stress when there is at least one diameter of the stud engagement into the cast iron. Rolled threaded bolts of fatigue resistant material are used. A program to keep them tight should in place, usually monthly and lengthen from there if appropriate. Some have suggested using Belleville washers to keep the studs from loosening.

3Q. We think the foundation block is cracked under a 2000 horsepower gas engine. How can we be sure and what can we do?

3A. Why do you think it is cracked? Vertical cracks are not uncommon. In one compressor house that had 20 gas engine compressors over 60 years old, just about all of them were cracked. You couldn't tell unless the engine frame was removed. This didn't cause problems because they were short base engines of 600 horsepower each. On one long base engine (around 30 feet) some of the block was above grade and you could see the crack "winking", which means opening and closing as the engine fired. It was just a few thousands of an inch. It caused much distress with bearing failures. Finally we used chocks and kept knocking them out as they deteriorated from the motion. The engine ran four years like this until it was replaced with a centrifugal compressor.

To verify you can have a crack you can have core samples taken, and map the crack. However you now have put more holes in the concrete block which itself could cause problems. There are contractors who can repair certain type cracks but they must be experienced. Tie bolts through the block to pull the halves together takes good engineering and experience. Done wrong and you could buckle the concrete block. Oil gets into cracks and degrades the concrete making it friable (crumbles). When pulled together too much, bad things can happen.

4Q. Our high horsepower gas engine compressor keeps beating out its bearings. What can be done?

4A. Babbitt type bearings are very forgiving with debris or slight misalignments. They don't care for heavy pounding type overloads as can occur with detonations. The babbitt layers can fatigue and pockets of material spall out from the surface. There are many different bearing designs which may be better for your operation and you should contact the manufacturer of the engine or a reputable bearing manufacture. Solid aluminum bearings are sometimes used for very heavy loading. Everything has to be right like the metallurgy, crush, clearance, etc. and the oil has to be super clean. When it is wrong you will be trying to remove aluminum that has become molten from the crankshaft and inside the engine case. Changing bearing designs is a job for the experienced. Obviously eliminating detonation and overloads might solve the problem with the present bearings.

5Q. On our 2,700 HP integral gas engine compressor we hear a metallic knocking sound during operation, which wasn't heard before. Any idea what it could be?

5A. Knocks are hard to analyze by mail. My experience with knocks have been that those coming from the engine frame, meaning badly worn bearings can sometimes sound like they come from the compression cylinder. I usually used a stethoscope type device to try and isolate the sound. Broke valves and loose slippers usually have a clicking type sound. Head clearance problems, excessively worn rider bands or worn steps in the cylinders, loose piston rod nuts, liquid slugging are more like knocks.

6Q. Is piston rod run-out and a bent piston rod shaft mean the same thing?

6A. No not really. Run-out is a method for measuring the horizontal and vertical motion of the piston rod within the engine and is a measure of wear and alignment of the cylinders on the compressor. A dial indicator or proximity gage is mounted to a fixed part of the frame and the vertical and horizontal motion of the rod is measured as the engine is rotated through a cycle. The limits are provided by the manufacturer instructions and depend on the cylinder running clearances. For piston to cylinder clearances in 0.010 inch range, run-outs in the 0.005 inch range aren't unusual, but it all depends on the type compressor. Bent shafts are measured out of the compressor and between centers, with the piston removed. Shafts should not run-out more than 0.001 inch. Over-tightening the piston rod nut has been known to kink (bend) the shaft.

7Q. Our compressor valve cover ears keep breaking off? What can we do?

7A. A nasty problem that usually can occur on older cast iron valve covers. Machinists tighten them up and because of the brittle nature of cast iron, it's easy to over-tighten and hear that unwanted "pop". Sometimes if the valve hold down bolt comes loose the valve can rotate in the cylinder on some compressors and impact the cap. This causes a rather spectacular release of product. On some engine we converted to cast steel covers.

8Q. We get heavy power knocking of our gas engine compressor in the summer. It's cracked frames and beaten out bearings. 800 horsepower engines are turbocharged. Any suggestions?

8A. The problem was that the engines did not have turbocharger air coolers on these old engines and the inlet air to the cylinders on the two cycle engines was too high. Adding air coolers solved the problem. Detonation caused such heavy loads that all the cast iron frames were badly cracked and had been stitched and bolted back together. All were replaced with used frames from a shut down plant.

9Q. We had a crankshaft failure on a large reciprocating engine? What can be the cause?

9A. Here’s a case where someone doing this investigation will need a lot of data. Photos of the failure, repair and overhaul history, historical failures, metallurgical report on failure area, up-ratings, operating conditions at time of failure and a host of other things.
Crankshaft failures were a rarity when I was working for a large diesel engine builder and doing design work. Later I witnessed three. One was on a gas engine compressor and was due to a poorly aligned tail bearing on an external flywheel. The second was on a large gas engine compressor and was due to excessive crank web deflection because of a degradation of frame grouting. The third was due to a propeller strike on an aircraft engine with a subsequent long-term fatigue failure. I also witnessed torsional failures of crankshafts due to locked up torsional dampener’s.

10Q. How do you determine if a crankshaft in a large engine is acceptable?

10A. There are many ways trained technicians can determine this. One method requires disassembly during an over-haul and examining the main and rod bearing wear. An experienced person will recognize wear patterns that can indicate a distorted crankshaft and other problems.

A method used without engine disassembly is to record the web deflection, or “pinching” in and out of the crank throws with a specialize dial indicator. This is done for each web at TDC, 90, 180, 270 and TDC. These readings are then used in a computer program to graph out the crankshaft distorted shape.

As you can see these methods require extensive experience in both utilizing the techniques and interpreting the results.