From: ahk@chinet.com   
      
   The disaster in Lac Magentic was caused by train MMA-002, operated by an   
   engineer without conductor, on Montreal, Maine, & Atlantic Railway running   
   away after being left unattended in Nantes. The train consist was 72 tank   
   cars and a buffer car (which protects the engine crew from disaster caused   
   by dangerous goods). There were five locomotives in the engine consist,   
   with a VB car (special-purpose caboose) behind the lead locomotive with   
   equipment for locomotive remote control. The engineer applied hand brakes   
   on 7 cars in total and shut down trailing locomotives, including 2 equipped   
   with auto-start that are not required to be shut down for fuel conservation   
   purposes as they shut off on their own eventually. The hand brake test was   
   performed without releasing the independent brakes (parking brakes) on the   
   locomotives, but this was a downgrade and it's not entirely clear if this   
   was a rules violation. Nonetheless, with independent brakes applied, it   
   didn't produce an adequate test of hand brakes.   
      
   Point 16, if I'm understanding it correctly, indicates why the engineer   
   thought the hand brakes were holding the consist after he performed the   
   roll test. It's clear that despite the engineer's long experience, he   
   was never trained adequately in calculating the number of hand brakes to   
   set nor what constitutes an adequate test of hand brakes. On page 127,   
   the point is made that MM&A's use of single person train operation   
   implementation didn't include training in adequate train securement. The   
   engineer must perform the conductor's duties as well as his own, and   
   apparently securing the train falls more heavily on the conductor portion   
   of this job. There was no mention if he first hired on at CP as a conductor.   
      
   A few points of interest from the Lac Magentic report:   
      
   1) The fire department shut down of the locomotive was done under railroad   
   direction, a fact I'm very glad to hear. Never read that in any of the   
   local or foreign stories I read of the disaster. Both the fuel line   
   valve was closed, to stop the fire, and the electrical breakers were   
   opened, to prevent a spark from igniting fuel. Unfortunately, MMA sent   
   a track foreman who wouldn't have a locomotive operating background and   
   didn't tell him to start another locomotive to keep the compressor going   
   to maintain air pressure in the air brake system. Normally, a penalty   
   brake application (dumping the air in all brake cylinders in the entire   
   train) would have occurred, except that the reset safety control (RSC)   
   was improperly wired so that this would not happen.   
      
   2) The engineer was rules qualified and not violating hours of service.   
   Montreal, Maine & Atlantic Railway is ex-CP trackage; MM&A is the   
   second short line to own it. The engineer had hired on to CP in 1980,   
   so he had plenty of experience.   
      
   3) Railroad industry assumes train crews are capable of exerting 125 lbs   
   of force on the brake wheel, but tests reveal that 80 to 100 lbs of force   
   is more typical.   
      
   4) Applying hand brakes on the locomotive is not assumed to be capable   
   of holding in place any additional cars. A new FRA standard applicable to   
   locomotives placed in service after January 4, 2004, requires the hand   
   brake to be capable of holding the locomotive in place on a 3% grade, a   
   net braking ratio of 10%. Generally, locomotive hand brakes met this   
   standard anyway prior to the effective date.   
      
   5) Applying the locomotive's hand brakes doesn't typically apply all   
   brake shoes. 2 of 12 brake shoes is typical, the way this model GE locomotive   
   was designed.   
      
   6) MM&A, as former CP territory, continued to use CP's General Operating   
   Instructions. It also refers to MM&A's General Special Instructions (GSIs)   
   and Safety Rules, without clarifying that these are additional documents   
   to those taken from CP.   
      
   7) Applicable number of handbrakes to apply uses a 10% of the number of   
   cars in the consist (72 in this case) plus 2. Setting hand brakes on the   
   locomotives counts toward the minimum. Hand brakes are required to be set   
   on the locomotive. Under this formula, 9 were to be set.   
      
   8) The grade at Nantes, where the train was parked, is .92%. From the   
   summary of rules given, assuming the consist holds for the hand brake test,   
   it doesn't appear that there was a requirement to set more than 9. However,   
   with 13 PSI automatic brake application to stop the train, 15 to 20   
   hand brakes should have been set, and without any air brakes, 18 to 26   
   air brakes should have been set. Interestingly, because of the wear on   
   the brake shoes on the locomotives' brakes, the engineer would have needed   
   to set between 12 and 18 brakes if just done on the cars.   
      
   9) Emergency brakes are applied with a complete dump of air. However, when   
   the air bleeds off slowly, and gets below 40 PSI, an emergency brake   
   application is no longer possible.   
      
   10) It took about an hour before the brake pressure fell to the point at which   
   the independent brakes (parking brakes) in the locomotive no longer held.   
      
   11) The sense and braking unit (SBU), a device connected to the rear of   
   the train attached to the brake line that can apply emergency brakes,   
   wasn't sufficient to apply emergency brakes in trains longer than 5   
   cars. By the time the device noted that the train had started to move,   
   brake pressure had dropped to 29 PSI.   
      
   12) In a test train with engine shut off, it took 1 hr 35 minutes before   
   air pressure dropped to 27 PSI, which is the pressure at which the train   
   started to roll. But the train's locomotive took 1 hr 6 minutes to drop   
   pressure to 27 PSI. Not ideal but not outrageous given the age of the parts.   
      
   13) The quick release brake (QRB) valve on the second locomotive in the   
   four locomotive engine consist failed. Typically, on a locomotive so   
   equipped, it opens to dump the air in the cylinder when the hand brake   
   is tightened when opened with the brake chain. The valve was worn and   
   damaged and, er, repaired in a non- standard manner (not explained). Unless   
   the air in the cylinder is dumped, the shoe isn't applied with the hand   
   brake. MM&A issued an instruction specific to these locomotives that the   
   crew member must listen for the air to be exhausted, otherwise the valve   
   must be operated manually, but the engineer wasn't aware of the memo.   
      
   14) Brake shoe testing of the locomotives indicated that some were worn to   
   through the lining to the backing plate. They can tell from wheel blueing   
   and lining wear that the independent brakes (parking brakes) had been   
   applying most of the braking force for the train. Finally, not all the   
   wheels showed full tread blueing nor brake shoe lining wear, indicating   
   that the hand brakes had not been, or could not be, applied securely.   
      
   15) 146,700 pounds retarding brake force was required to keep the   
   consist parked. The calculations are on page 27 if you care to read them.   
      
   16) Hand brakes are to be applied after full release of air brakes,   
   except that it's not possible to fully release air brakes when the train   
   is stopped on a grade. So the 13 PSI automatic brake application at   
   Nantes (where the consist was parked) resulted in hand brake forces   
   40% higher than without air brakes. Page 38   
      
   17) Reset safety controls (RSCs) are features of locomotives manufactured   
   since 1986. The 3 pre-1986 GE locomotives were retrofitted with them   
   by a previous owner. RSCs incorporate the dead man's switch feature,   
   but also apply a penalty brake application in the event of opening the   
   breaker or shutting off the main electrical power. There's no requirement   
   that the penalty brake application function work with power loss, so   
   it's not required that this be tested for in the shop. Wiring of this   
   device was inconsistent among the three locomotives and the penalty brake   
   application didn't necessarily occur as it was supposed to. In fact, 5   
   other GE locomotives owned by MM&A had inconsistent wiring problems. In   
   one of the locomotives in the engine consist, RSC was wired directly to   
   the battery and therefore remained powered even though the main electrical   
   power was shut off.   
      
   Now, penalty brake application dumps the air in the cylinder, but I don't   
   understand how long the brake shoe is held in place. It's not an   
   indefinite application of brake force like a properly set hand brake, right?   
      
   18) Pages 29-31 and 113 discuss the serious problems with the lead   
   locomotive and temporary and improper maintenance peformed on it. The   
   temporary repairs were explained by MM&A having no spare locomotives and   
   an increase in traffic due to oil trains. Gee, you'd think they could   
   have rented one so they'd have had time to perform proper maintenance. The   
   locomotive remained in service despite problems being reported. The most   
   serious maintenance issue was the shop "repairing" an oil leak at a cam   
   bearing bore by overtightening the mounting bolt. After the disaster,   
   testing revealed that the bearing had fractured.   
      
   --- SoupGate/W32 v1.03   
    * Origin: LiveWire BBS -=*=- UseNet FTN Gateway (1:2320/1)