Investigating a fuel issue
How extensive troubleshooting lead to the replacement of a fuel tank from a sister loco
50049 Defiance had earlier this year been experiencing issues of low power, which in a previous post had been attributed to a failed turbocharger unit. This failed unit was duly swapped and we were back to a fully operational capacity, however all was not well as 50049 was still struggling with low power. After spending any length of time with something, whether this be a car, motorbike, or in this instance a Class 50, you start to pick up on the quirks, sounds, and noises of each loco and you understand what is “normal” and this forms the basis of when things aren’t quite right.
With 50049 not being “quite right”, it was time to start fault finding into the root cause of this problem as it could be a whole host of items, so it was time to start working through the list bit by bit…
The first items to be inspected and changed were the fuel filter and strainer, these are two items that are easily accessible. Both items were checked and were absolutely fine but were changed and cleaned as a matter of course. Onto the next item…
The Woodward governor is often referred to a being “The Boss” as this is the key component which controls the amount of fuel being sent into the combustion chamber of the power unit by opening and closing the fuel racks depending on power demand, and the current load. There are a multitude of differing variables which are all key inputs to the governor and if you have ever seen one with the covers off, they are more akin to a top of the range Swiss Watch with how intricate the mechanisms are. This was the first item to be swapped. As 50033’s power unit had been removed for the main generator to be removed, along with a full overhaul of the engine room, the governor from 50033 was swapped onto 50049 and this was trialled. 50049 was taken out for a test and the low power and bogging down of the power unit still existed. Onto the next item…
Class 50’s are not fitted with fuel pressure gauges, however we did have one which was often used on our loco’s when any major power unit overhauls were completed, at the time it was fitted to 50044, so this was removed and placed on 50049 for more testing to be undertaken. She was fired up and left to warm through with the pressures being constantly monitored, and they were what we would expect to see. With the idle pressures being correct, she was taken for a test run to see how the pressures would fair with the power unit under load and requiring more fuel supply. After the first 2 runs, all was looking well, however then on the final run, the pressure dropped and the power unit could be heard bogging down, this was due to fuel starvation. At this point we had isolated what the issue was, now it was going to be a case of delving into the fuel system to understand the root cause.
The fuel lift pump
The fuel lift pump
We were going to go through the list in order of least complexity to trial and change first. First up was the pressure relief valve, no change. Next up, the dampening vessel, no change. Then came the larger of the items… the fuel lift pump, this lifts the fuel from the tank via vacuum and then pumps this through to the individual fuel pumps to deliver to the injectors. As 50033 didn’t have its power unit in place, this was the ideal candidate to trial the fuel lift pump as it is mounted on the No1 end bulkhead of the engine room. This was removed prior to removing the item in 50049. Once removed from 50033, the coolant was drained from 50049 as several coolant pipes need to be removed in order to facilitate the removal of the fuel lift pump. Once this was completed the pump was disconnect and removed and 50033’s item refitted and reconnected.
After pumping the coolant back into 50049 she was fired up and all pressures looked great. She was taken for a test run and all of the pressures looked fine. An ECS move was required by the SVR so with all looking promising 50049 was drafted in. The railway was closed to the public at this point so if any problems arose it wasn’t going to affect any public trains and was a controlled test with the ability for another loco to rescue if required. 50049 performed the move with success but whilst the power unit hadn’t bogged down, it was noted that the pressures just weren’t sufficient enough. Now for short spells at maximum power demand on the SVR this would probably suffice, however for 50049 to operate at sustained maximum power demands on the mainline this was not sufficient in any way, shape or form. At this point we had changed everything within the fuel system aside from one item… the fuel tank.
The fuel tank on a class 50 is a 1050-gallon item so isn’t a small item by any means, however could it be the cause of this low fuel pressure….? Prior to committing to dropping the fuel tank we wanted to complete a trial before hand.
The trial consisted of disconnecting the fuel lift pump from the main fuel tank, and connecting it to a temporary fuel supply using the same diameter pipe/hose. We managed to create a temporary fuel supply which supplied enough fuel for us to complete a short test within Kidderminster Station limits. This was the test which would determine whether we dropped the tank or not. Maximum power was demanded by the driver and low and behold 50049 took off like a scalded cat! It was a “eureka” moment, we’d finally reached the final item of the list and through the testing of the temporary supply we’d proved it!
Now came the task of dropping the fuel tank and trying to understand what was the root cause…. As time was against us due to a rail tour commitment, the decision was made to remove 50031’s fuel tank and then to swap this beneath 50049. Removing one fuel tank was difficult enough as this had been completed some years ago on 50044 in order to facilitate a repair on one of the air pipes beneath the loco, however the team were going to embark on the removal of two fuel tanks and the refitting of one in a weekend.
Prior to the removal of both fuel tanks, the fuel required emptying into the rest of the fleet which was duly completed. Both locos entered into the TMD, the tank would be removed from 50031 first so that if any issues were encountered, we weren’t going to isolate two locos, and would only isolate the one loco currently stopped from active use.
The tank was dropped onto pump trucks and slid out of the way. 50049 was next up, and the tank was removed without issue using the same method and set to one side. Prior to 50031’s tank being put beneath 50049, it allowed us time to inspect all of the pipes beneath 50049, and to also give 50031’s a thorough cleaning prior to refitting. This level of activity would’ve only been completed at Works. Possibly on a major overhaul, and if any other issues existed, but it was very rare. After we were happy with the condition of the tank, this was move into place and fastened into place, with all fixings torqued correctly and systematically.
Once everything had been check, checked and double-checked, fuel was put into the tank and 50049 was taken for a test. This was the moment of truth, had all of the hard work paid off? In short, yes it had! 50049 passed with flying colours and fuel pressure was back!
With the fuel pressure back to being normal, there was still the problem that 50049 still had 50033’s governor on the power unit which was required back. This was removed and put back onto 50033 prior to 50033 being return back to service, and due to a seal having potentially failed on 50049’s governor causing oil contamination, it was decided to put on a spare governor which was from 50036’s power unit.
Whilst we were original swapping governors for the trial it was noted that one bank on 50049's power unit was running richer than the other, so all fuel racks were completely reset and all fuel pumps reset back to a nominal and tuned and adjusted into the governor to allow for maximum efficiency both at maximum power demand and idle. Many of you may have noticed that 50049 used to have a slight “chug” on idle, you may have noticed that this has now been completely eradicated after overhauling and tuning all of the settings.
Inspecting the removed tank
Inspecting the removed tank
After returning 50049 successfully back to traffic her fuel tank was inspected to find out what the root cause of the fuel starvation was. The clamps that held the suction pipe within the tank had slightly loosened over the period of time allowing the pipe to drop slightly, and sludge was found to be at the bottom of the tank which had accumulated over the life of the fuel tank. The unfortunate coupling of both issues was the recipe for not enough fuel being able to be sucked up the pipe and to the fuel lift pump due to the partial blockage. The remedy for this will be the re-securing of the suction pipe, and to give the fuel tank a thorough clean. Now that we have found this issue on one of our locos, we will systematically complete this planned maintenance task on the rest of the fleet.
Ready for traffic
Ready for traffic
The above has been a huge team effort over several months, but we were never going to give up until having found the root cause, and solving it.
A huge thanks to everyone who was involved in this task as it was simply remarkable!
Pictures courtesy of:
Class 50 Alliance