• Boiler explosions?

  • Discussion of steam locomotives from all manufacturers and railroads
Discussion of steam locomotives from all manufacturers and railroads

Moderators: Typewriters, slide rules

  by Rail-Gun
 
Does anyone know where i could possibly find a photo the hoffific boiler explosion of SP 4-10-2 No.5037 and of the
C&0 2-6-6-6 No.1622. I know these are horrible and tragic events but these are intresting how these seem to be the only rather large super heated locmotives that i can find that suffered boiler explosions. Ive been researching how steam engines work and i came across a few chapters of boiler explosions, it was rather shocking and surprising to me to see how catastrophic a boiler explosion can be. Thank you for your time
  by slide rules
 
The book "The Allegheny: Lima's Finest" by Eugene Huddleston and Thomas Dixon Jr. has quite a few pictures of the 2-6-6-6 after the boiler explosion, and an excellent explanation of what they think happened and why. This is the only place I have ever seen pictures of that explosion, as for the SP 4-10-2 I will have to get back to you on that! I dont know if the Allegheny book is still available or not, but if it can be found it is an excellent book, and a fine addition to any locomotion library.

David A. Davis
  by wmmanager
 
I have the final ICC report on the 1622 event I can email you if you like.

Or if it's ok with the moderation, I can paste the report here.

Loyd L.
  by Typewriters
 
Since the ICC report is not of a restricted nature, i.e. is a matter of public record I see no problems with posting it here.

-Will Davis
  by wmmanager
 
Here you go guys. The entire final report posted word for word.

Loyd L.

INTERSTATE COMMERCE COMMISSION
WASHINGTON
REPORT NO. 3520
CHESAPEAKE & OHIO RAILWAY COMPANY
IN RE ACCIDENT AT HINTON, W. VA., ON JUNE 9, 1953
Report No. 3520
SUMMARY
Date: June 9, 1953
Railroad: Chesapeake & Ohio
Location: Hinton, W.Va.
Kind of accident: Boiler explosion
Train involved: Freight
Train number: Extra 1642 East
Locomotive number: 1642
Consist: 123 loaded and 2 empty cars
Speed: 20 m.p.h.
Operation: Freight movement
Track: Level and tangent
Time: 5:25 p.m.
Casualties: 3 killed
Cause: Overheated crown sheet resulting from low water


REPORT NO. 3520


IN THE MATTER OF MAKING ACCIDENT INVESTIGATION REPORTS UNDER THE LOCOMOTIVE INSPECTION ACT OF FEBRUARY 17, 1911, AS AMENDED


CHESAPEAKE & OHIO RAILWAY


July 6, 1953


Accident (boiler explosion) at Hinton, W. Va., on June 9, 1953, caused by overheated crown sheet due to low water.


REPORT OF THE COMMISSION 1


PATTERSON, Commissioner:


On June 9, 1953, about 5:25 p.m., at Hinton, W. Va., the boiler of Chesapeake & Ohio Railway locomotive 1642 exploded while the locomotive was hauling a freight train at an estimated speed of 20 miles per hour. The engineer, fireman, and head brakeman were killed.


DESCRIPTION OF ACCIDENT


Chesapeake & Ohio Railway locomotive No. 1642, hauling eastbound freight train Extra 1642 East, departed from Handley, W, Va., at 1:20 p.m., June 9, 1953, and proceeded without any known unusual incident to CW Cabin, near the city limit of Hinton, W. Va., a distance of 71.6 miles from Handley, where, about 5:25 p.m., the boiler of the locomotive exploded while the train was moving at an estimated speed of 20 miles per hour.


The train left Handley, W. Va., with 91 loaded, cars, adjusted tonnage 7510 tons. A stop was made at Thurmond, W. Va., 38.6 miles from Handley, at 2:55 p.m., where coal and water were taken, and a stop was made at Quinnimont, W. Va., 12 miles from Thurmond, at 4:12 p.m., where water was taken and cars were picked up. The train departed from Quinnimont, approximately 21 miles from the point of the accident, at 4:38 p.m. with consist of 123 loaded and 2 empty cars, 10,430 adjusted tons. The tonnage rating for the locomotive over this part of the division was 11,500 adjusted tons. Approaching the scene of the accident the track was undulating, but at the point of the explosion was level and. tangent. The weather was clear and dry. The positions of the three employees on the locomotive at time of the accident were not known.


At the point of the explosion, there were two tracks on the left side of the eastbound main, the westbound main and a switching lead, and on the right side New River ran approximately parallel with and about 55 feet from the eastbound main.


The force of the explosion tore the boiler from the frame and cylinder connections and it was thrown upward and forward. The boiler struck on its front end on the rails of the eastbound track approximately 440 feet ahead of the point of the explosion, then rebounded. The back head, struck the track 639 feet ahead of the point of explosion where the boiler came to rest on its right side in reversed position with front end, on the adjacent westbound track and firebox on the switching track. The smoke box front was blown off and several super-heater units were blown out. The cab was blown 133 feet to rear and 58 feet to right of the point of explosion where it fell at the water edge of New River. Grates, grate bars, throttle lever, and other parts were scattered for distances up to approximately 772 feet from point of accident, some falling in New River, Many appurtenances were badly damaged and some parts could not be located. The track rails at point of explosion were indented by the trailing truck wheels and the two rear pairs of driving wheels and the westbound track was moved approximately 5-1/4. feet to the left. At the point where the front end of the boiler struck the track rails were broken and badly bent and a large hole was torn in the road bed. Where the back head of the boiler struck, the westbound track was moved 3 feet to the left. The locomotive running gear with tender attached came to rest with front end alongside the front end of the boiler with only trailing truck wheels derailed. All tender truck wheels were derailed and the front truck was off center. The tank was skewed, to the left with left front corner leaning approximately 10 degrees to the left. Nine cars were derailed and bunched, five were at approximately 90-degree angles with the rails four of which were on their sides.


The engineer, fireman, and head brakeman were killed. The engineer's body was found at the water's edge of New River, approximately 75 feet to rear of the cab. The fireman's body was found in the cab, and the brakeman's body was found in a ditch on the left side of the tracks near the point of the explosion.


DESCRIPTION OF LOCOMOTIVE


Locomotive 1642, 2-6-6-6 type, carrier's classification H-8 Alleghany, was built by the Lima Locomotive Works Inc., at Lima, Ohio, in December 1944. The four cylinders wore 22-1/2 x 33 inches the diameter of driving wheels 67 Inches with new tire's weight in working order 771,300 pounds, weight on driving wheels 507,900 pounds, and tractive effort 110,200 pounds. The locomotive was equipped with an Alco Typo H power reverse gear, American multiple front throttle, Standard H D stoker, Franklin No. 8-A Butterfly mechanically operated fire door, Baker valve gear, Worthington Type 6-1/2 S S A feed water pump, Nathan Type 4000-C special injector. The boiler was equipped with a Nathan Type B low water alarm and there were three Nicholson theremic syphone in the firebox. Locomotive had made 97,000 miles since last Class 3 repairs and 18,000 miles since last class 5 repairs. The rectangular east-steel water-bottom tenier had capacity of 25,000 gallons of water and 25 tons of coal.


The boiler was of the three-course conical type with combustion chamber and wide radial-stayed firebox; builder's serial boiler number 8811. The inside diameter of the first course was 101-1/8 inches, second course 103-11/16 inches, and third course 106-5/16 inches, second course 103-11/16 inches, and third course 106-5/16 inches; thickness of first course 1-9/32 inches, second course 1-5/16 inches, and third boiler had 48 2-1/4 inch outside diamtor flues and 278 3-1/2 inch outside diameter flues, 23 feet in length, and 71 Elesco Type E superheater Units. The working steam pressure of the boiler was 260 pounds per square inch.


The radial-stayed firebox was 180 inches long and 109 inches wide, and combustion chamber was 118 inches long. The firebox consisted of a one-piece crown and upper side sheets, lower one-fourth side sheets, door sheet, flue sheet, and inside throat sheet. Flue sheet and throat sheet were 9/16 inch thick and other' sheets were 3/8 inch thick, Flue sheet seam was riveted and door sheet seam was riveted across the top and welded down the sides. Other seams and patches in the firebox were butt welded. The crown sheet was 11-3/4 inches higher at the flue sheet than at the door sheet, The firebox was fitted with three thermic syphons. There was no syphon in the combustion chamber. New flue sheet and lower side sheets were applied on April 6, 1950,. at which time a patch was applied in bottom of combustion chamber, one-half section applied to left syphon, and patches applied to center syphon and, to diaphragm of connection sheet, Crown stays wore 1-1/8 inch diameter reduced body type, spaced approximately 4-1-16x4 inches Combustion chamber stays were 1 inch diameter, spaced approximately 4-1/6 x 14 inches. Fire box stays were 1 inch diameter, spaced approximately 4-1/8 x 14 inches. All stays were rigid except in the combustion chamber and breaking zones.


EXAMINATION OF BOILER AND APPURTENANCES BOILER


The crown sheet had boon overheated its entire width at flue sheet, the overheated area extending to the 12th row of stays on each side of longitudinal center at front end and tapering gradually upward and backward to the 1st row of stays on right and left sides of center syphon at the 57th transverse row. The line of demarcation was distinct and indicated the water had been approximately 7-1/4 inches below the highest part of the crown sheet. Grown sheet had evidently initially pulled from 123 stays and pocketed at the front. center, The stays in this pocketed area were a deep blue in color, stay ends were cupped to a maximum depth of 1/4 inch, and stay holes were elongated to a maximum diameter of 1-3/4 inches. The sheet was not thinned to any noticeable extent.


The back flue shoot ton through the top row of flues from the flue sheet flange at 11th row of crown stays on right side to the 13th row on loft side. The tear continued into the sides of combustion chamber, terminating in vicinity of the 20th longitudinal row and 5th transverse row of stays on each side, The top port of flue sheet below the tear was pulled from 41 flues and folded down, The brown sheet and side sheets above tears in the combustion chamber were blown down against the bottom of combustion chamber, the folds on each side starting at the ends of the tears in sides of combustion chamber shoot,


Irregular tears practically crossed the crown sheet between the 24th and 28th transverse rows of stays and extended down in the side sheets to about the 20th longitudinal rows of stays on each side, Other irregular tears crossed crown shoot at about the 39th transverse row of stays and extended down in side shoots to about the 20th longitudinal row of stays on each side, Irregular longitudinal tears joined, the ends of these transverse tears in the side sheets. The rear row of tears across crown sheet were just ahead of the thermic syphons. A large part of the torn out portion of crown sheet folded down over the throat sheet and loft syphon. The three syphons were pulled out of inner throat sheet; were badly bent and center syphon was broken through more than 50 percent of its cross-sectional area at the neck.


The crown sheet was pulled from approximately 861 stays. A total of 1587 stays were pulled from the crown sheet, combustion chamber, and side sheets, Threads on crown stays and in stay holes appeared to have been in good condition prior to the accident. There were no broken crown stays or stayblots in firebox sheets. No crown stays or staybolts showed any indication of having been worked excessively, and all flue ends appeared to have been in good condition previous to the accident. There was a slight amount of scale on the sheets.


The back head and the roof sheet and door sheet were dented when the boiler struck the track rails. Both sides of the mud ring were sprung outward 13 inches at the center.


APPURTENANCES


Safety valves: The boiler was equipped with four 3-1/2 inch Consolidated safety valves, three open and one muffled type, located on top of the third course. Safety valves were not badly damaged in the accident, but the right safety valve nipple was partially pulled from the boiler, The safety valves were applied to locomotive 1636, same class as locomotive 1642, and each valve tested twice A test gage was mounted adjacent to the safety valves and a certified gage was used in the cab. On both trials, No. 1 valve opened at 25 lbs and closed, at 252 lbs, No. 2 valve opened at 260 lbs. and closed at 252 lbs, No. 3 valve opened at 262 lbs, and closed at 256 lbs, and No. 4 valve opened at 266 lbs. and closed at 256 lbs. The No. 4, valve simmered at 262 lbs. and opened fully at 266 lbs.


Steam gage: An Ashoroft 400-pound 6-inch double-dial steam gage which had been mounted at center of boiler back head was not recovered Steam gage valve and siphon pipe wore found broken and twisted.


Water level indicating devices: The boiler was equipped with a Nathan 300-pound water column, located 29-l/2 inches to the right of vertical center line of boiler beck hoed, Three gage cocks and a 6-1/2 inch reflex water glass were applied to the column. An additional reflex water gloss was mounted on the left side of the boiler back head 26 inches left of back head vertical center line. The water column, both water glasses and all water-glass valves, and gage cocks were made by Nathan Manufacturing Company.


All water level indicating device connections to the boiler were broken off or torn out by impact.


The bottom connection between water column and boiler was located 16 inches above the horizontal center line of the back head. A 1-1/2 inch O. D, copper pipe extended from the top of the water column to a company's standard spud which was located 12 inches ahead of wrapper shoot calking edge and 6 inches to right of the top center line. The bottom water-column spud which extended into the water space 4 inches was crushed, but its 3/4-inch opening appeared to be unobstructed The top column steam pipe was destroyed and the wrapper shoot spud broken off. The 1-1/4 inch opening in the spud was found clean and unobstructed. The interior of the water column was free from scale and mud deposits; the 3/4-inch drain valve was torn off but was found in closed position and operated freely when tested.


Three double-seated gage cocks were spirally mounted on the water column with 3-inch differences in height. The gage cocks which wore broken off and damaged, could not be tested under pressure. The 5/16-inch opening into the column were unobstructed. Visual inspection of component parte when disassembled should no evidence of the lockage or cause. Carrier's records shared the lowest gage-cock opening had been 6-1/2 inches above the highest part of the crown sheet and level with the lowest reading of the water glasses.


Right water-glass valve connections wire broken off flush with the water column, leaving clean 3/8-inch holes. The 5/8 inch O.D. copper steam pipe connecting the right water glass to the water-glass valve was found it good condition


The left reflex water glass was mounted 26 inches to the left of vertical center line of boiler back head with the bot-torn connection 25 inches above the horizontal center line, A 5/8-inch O. D. copper steam pipe connected the water glass to a company's standard spud which entered the boiler 6 inches to the left of center line arid. 12-1/2 inches ahead of the wrapper sheet calking edge. The bottom connection to the left water glass had been broken off. The bottom spud extended 3 inches into the water, space and its passageway was unobstructed. The left top water-glass steam pipe and spud were not recovered.


The right and left top and bottom water-glass stop valves were found in fully opened position, The 3/8-inch valve openings were unobstructed. The threads on both water-glass bodies were damaged, but the unbroken reflex glasses showed a clear water line at all heights when tested with cold water. The drain valves from these valves were not recovered.


Injector: The Nathan Type 4000-C injector, which had capacity of 13,000 gallons per hour, remained attached to the right side of the main frame. Its steam pipe, delivery pipe, starting lever, and extension to overflow valve were torn off and the injector was found in badly damaged condition. Four company officials stated that the steam valve and regulating valve were found in closed position, A new overflow valve stem was applied in order that tests could be conducted on locomotive 1636. During a two-minute test with boiler pressure at 205 pounds, the water was raised 1-3/4 inches, The pressure was raised to 255 pounds; the level of water lowered to conform with original height, and approximately identical performance was obtained in a second two-minute test. Starting when the 266-pound safety valve lifted, the injector was tested at various stages of descending boiler pressure. These tests demonstrated the injector functioned properly until the steam pressure had fallen to 120 pounds.


The carrier's drawings showed the starting lever for the non-lifting injector was of the latched lever and, quadrant type, and secured to the floor at the left side of engineer's seat box, 24 inches ahead of the back wall of the cab.


Feed water pumps: The boiler was equipped with a Worthington Type 6-1/2 S S A feedwater pump, 14,400 gallons per hour capacity. The turbine-driven cold water pump with attached feed water hose was broken from the bracket at the left rear main frame extension. The strainer and its compartment were clean and the strainer was found in proper position. The pump could not be tested because of the damaged condition of the impeller housing and water discharge fitting. The governor steam control valve was removed from the cold water pump and tested on locomotive 1610 in the condition as found. Results of the tests indicated this control valve functioned practically identically with the original equipment of locomotive 1610.


The feedwater heater was so badly damaged that any previous leakage from the system and its related piping could not be demantled. The drifting control steam valve was dismantled and its spring and valve were found in good condition.


The hot water pump was broken through the center member and the piston rod was bent approximately 20 degrees. All parts of the steam portion of this pump, including reversing valve, were well lubricated and worked freely; packing ring and valve ring fit and pressure against the cylinder walls were good. The hot water portion of this pump was also found in good condition. All twelve wing-type valves were found seated and valve springs had good resilience, Valves had good contact with the seats. There were no foreign objects found in the cylinders or pump passages.


The manifold steam valve end piping, with throttle valve attached had been separated from the manifold. The manifold valve was found in open position. The 1-1/2 inch 300-pound Lukenholmer throttle valve was found completely closed with threaded valve stem bent.


The hot and cold water pumps, drifting control valve and governor control valves were disconnected after the accident and examined. Visual inspection did not indicate any defective conditions.


Boiler checks and delivery pipes: The 3-1/2 inch delivery pipes were badly damaged but the check valves and stop valves remained attached to the boiler. The stop valves were found in open position and were clean. The 3-inch right boiler check valve, located on the first boiler course, operated freely and had 1/16 inch lift in excess of the carrier's standard. This valve body had a small deposit of soft scale. The valve and its seat were in good condition. A corresponding check valve, located above the center line on the left side of the first course had lift 5/32 inch in excess of the carrier's standard and was found clean. The valve and, its seat were in good condition and the valve was free.


Blow-off cocks: The boiler was fitted with four 2-inch Okadee blow-off cocks located near the mud ring corners. The two front blow-off cocks were piped to a blow-down separator located on top of the boiler and manually operated from the right and left sides of the cab. The right back blow-off cock was torn off and its valve was found seated and could not he operated manually. The other three cocks remained attached, but due to damage of the operating mechanisms, the former valve positions could not be determined.


Low water alarm: The exterior parts of the Nathan Type B low water alarm which had been located on the third course of the boiler were damaged and the interior drop pipe was twisted from normal position. The cab alarm whistle and pipe wore found crushed against the boiler back head. Carrier's records, dated March 29, 1950, indicated that the water level at which the alarm would function was 6-3/14 inches above the highest point of the crown sheet.


Feedwater tank, tank valves, hose and strainers: The feedwater tank valves were found fully open, and the 14-1/2 inch feedwater hose remained attached to the right side. The 8-inch circular copper strainer in the feedwater line to the injector was not found, The left hose was found with the cold feedwater pump, Both tank hose were in good condition. There was between 3/14 and 1 inch of scale and rust flakes in the bottom of the tank which could have been dislodged by shock at time of the explosion, The carrier's standard water level gage was observed by first witnesses and showed water at the second opening approximately 21 inches from the bottom of the tank.


Boiler water condition: Records of boiler water hardness on file at Hinton, W, Va., for June 8, 1953, showed 90 grains in-bound and 85 grains outbound, On arrival at Handley, W. Va., on June 9, 1953, the hardness was shown at 70 grains, and when last dispatched from Handley the reading was 50 grains.


INSPECTION AND REPAIR REPORTS


The last annual inspection was made at Clifton Forgo, Va., on July 25, 1952. The last monthly inspection was made at Hinton, W. Va., on June 5, 1953. The locomotive was out of service for the months of February, March, April, and May, 1953, and was returned to service on June 5, 1953. Daily inspection and repair reports since that date, from all points from which the locomotive had been operated, were examined and the following items which might have any bearing on the accident wore found reported:


June 5, Engineer reported: Water pump not working, clean governor,

Item signed for and report approved.


June 5, Machinist reported: Clean cold water pump governor.

Item signed for and report approved.


June 5, Engineer reported: Water pump quits working, clean and examine governor.

Report approved.


June 7, Engineer reported: Water pump stopped 3 or 4 times.

Item signed: Tested OK. Report approved.


June 8, Engineer reported: Water pump wont work, renew governor in cold water pump.

Item signed: Tested l-l/14" per min.

Report approved.


At all terminals on the Chesapeake & Ohio Railway, the feed-water pumps and, low water alarms are tested outbound. On June 9, at 11:30 a.m. at Handley, W. Va., a machinist filled out the prescribed form showing these tests had boon made.


SUMMARY OF EVIDENCE


The engineer who operated locomotive 1642 on its next to last previous trip and who was the last engineer to handle the locomotive on the road prior to the engineer who was killed in the accident stated that nothing unusual occurred on that trip and that the feedwater pump and injector operated satisfactorily. His fireman on that trip also stated that no trouble was experienced during the trip; that ho operated the water pump without difficulty; that the injector was also used and functioned properly.


The foreman at Thurmond, W. Va., stated that when the locomotive took coal at that point on the trip on which the accident occurred the engineer asked him to look at the cold water pump governor and see if it was stuck; it was examined, found free, and put back in. He then went with the engineer into the cab to examine the squirt hose which operated from the cold water line, the cab was washed down, and the locomotive was put back on the train.


A machinist at Handley, W. Va., the point from which the locomotive was last dispatched, stated that he tested the water pump and it raised the water level line 1-1/14 inches per minute and that he did not find anything wrong with the pump.


A machinist helper, who was between 150 and 175 feet from the track and, approximately 1-1/2 miles from the point of the explosion, stated that when the locomotive passed by him the engineer was seated in his usual position in the cab; the fireman was in a bent position on the left of the engineer; that the low water alarm whistle was sounding, and the exhaust from the stack sounded as if the engineer was working a medium throttle.


A roundhouse foreman and a sheet metal worker stated that they arrived at the scene of the accident about 5:30 p.m. and saw water running from the left tank hose which had been severed; that it continued to run until about 6:10 p.m., and that no water was coming from the injector overflow.


The telegraph operator on duty at CW Cabin at the time of the accident stated that he received a telephone inquiry Concerning location of Extra 1642 East; he looked down the track and saw the train approaching from a distance of about 600 foot; he arose and as he again looked at the approaching train trio explosion occurred. He stated the locomotive appeared to disintegrate, then was obscured by steam and smoke. After parts of the locomotive stopped falling, he called the train dispatcher and reported that locomotive 1642 had blown up and was wrecked in front of the office and requested that an ambulance be called. He further stated that he noted nothing unusual when he first observed the train approaching and that the locomotive sounded as though the engineer was working a medium throttle.


CAUSE OF ACCIDENT


It is found that this accident was caused by an overheated crown shoot due to low water,


Dated, at Washington, D. C., this 6th day of July, 1953.


By the Commission, Commissioner Patterson.


SEAL GEORGE W. LAIRD,


Acting Secretary.


FOOTNOTE:


1. Under authority of section 17 (2) of the Interstate Commerce Act the above-entitled proceeding was referred by the Commission to Commissioner Patterson for consideration and disposition.
  by slide rules
 
Thanks so much for posting the complete report! Very interesting reading, and a few things stand out for me after going over this report, along with the account given in print already mentioned.

An employee stated that moments before the explosion, he clearly heard the low water alarm whistle sounding as the locomotive passed him by. This alarm sounds when the water in the boiler is about 6 inches or so (see report for exact setting) ABOVE the crown sheet, yet inspection of the damaged boiler showed evidence of the water level having been allowed to get down to 7+ inches BELOW the level of the crown sheet. Now, the crew HAD to know what was going on, that the water was low, and working a light to medium throttle it isnt like steam usage was at its maximum rate. I have always wondered what the crew did as the alarm sounded, and continued to sound.......

It is safe to presume that they were using only the Worthington feedwater heater system to supply water to the boiler for the entire duration of the trip. This would not have been at all unusual. When the low water alarm sounded, even if they had not noticed the water level in the water columns dropping, the first thing they would have done is to reduce throttle, and check the feedwater heater pump (hot and cold) control settings. At the rate they were moving, the feedwater hot and cold pumps would not have been operating at anything like max capacity, so they (the fireman) should have "opened it up" and increased the rate at which water was being pumped into the boiler, while the engineer reduced throttle to slow steam usage. In fact, the witnesses stated he was working a MEDUIM throttle, surprising if he realized the severity of the low water condition. If the engineer believed that he could not get water into the boiler at a rate exceeding usage, he would have stopped the train, or SHOULD have. Did he not believe the level indicating devices? Did he think that the water might indeed be low, but controllable? Did he think that the margin of 6 or so inches of water above the crown sheet had not been used up, when in fact the water was somewhere around 13 inches BELOW the alarm level?

Mechanical inspections indicate that there had been repeated problems with the cold water pump in the feedwater heater system. Plus, earlier in this very trip there had been sht suspicion of a problem with the governor on the cold water pump. Assuming that there WAS another problem with it, and for whatever reason it could not be controlled or perhaps not run at sufficient speed to feed enough water (the governor could have again been at fault, so its hard to say WHAT the pump might have been doing, or been able to do) the next thing to do would have been to utilize the Nathan injector, whose controls were on the right, or engineer's side of the cab.

All accounts indicate that the controls for the Nathan injector were closed, and of course not running at the time of the explosion. The witness stated that the fireman was seen next to the engineer, in a bent position. Was he preparing to use the injector, having given up on the feedwater heater? Did the engineer have him come over to start it while he watched the road? What else could he have been doing? Discussing the problem with the feedwater heater? We will never know. But, the Nathan was never opened. Either device, operating nominally, had enough capacity to feed the boiler up to a very high rate of water usage.

A boiler equipped with Nicholson thermic syphons could (and would) last longer under low water conditions as compared against one without, as the lifting action would continue to keep some water playing over the crown sheet even if the level were low. This was one of the advertised benefits of the syphons. In this case, we can assume that they bought the crew extra time as the water level fell. But we still come back to asking just how long the low water alarm had been going off, and why wasn't effective action taken to prevent this explosion? As stated, they had to know about it, and had to have been operating with low water for some time.

It would seem to me that to use a more modern term usually seen in aviation disaster reports, this explosion would have to be chalked up as "pilot error", or more exactly crew error. No matter what the boiler feed devices were doing, low water was a serious, dangerous, immediate problem, and the crew HAD to have known that. If in fact the feedwater heater's cold water pump was acting up again (as it had in the past), the right hand injector should have been immediately utilized, or the train stopped until the problem could be sorted out. With the train stopped, inspection of the cold water pump would have been easy. If there were trouble with the right hand injector, for instance if they had in fact TRIED to use it but it was inoperable, stopping would have been the only safe way to proceed. For whatever reason, they did not stop, and operated with dangerously low water for what must have been some time. I have offered some possibilities, and a LOT more questions, but in the end, like similar disasters, we will never know for sure.

David A. Davis
  by Allen Hazen
 
Thank you, David Davis, for the intelligent comments. The locomotive crew's behavior is puzzling: if the low-water alarm was functioning normally, and the water level had dropped more than a foot from the level it was set for, they must have been listening to the alarm for ?? several minutes ?? at least. (Do low water alarms frequently sound "false positives"? If so, perhaps locomotive crews were tempted to break the rules and ignore them.)

Noteworthy that feedwater pump problems had been noted more than once in the week prior to the accident. Suggests a mechanical problem that (i) showed itself intermittently (ii) hadn't been fully repaired, so could recur when (to make up a conceivable cause just for the sake of example) vibration moved a chunk of rust inside the pump.

Just before the concluding summary we are tole "At all terminals on the Chesapeake & Ohio Railway, the feed-water pumps and, low water alarms are tested outbound." Water pump test is meaningless if the mechanical problem was only intermittent. And how do you test a low water alarm?

One thing struck me as dubious. Eye-witness testimony is notoriously unreliable. The machinist helper (who reported hearing the low-water alarm whistle) claimed to have seen a lot in the cab of a locomotive passing him (at a distance equivalent to most of a city block) at 20 mph.
  by O-6-O
 
A D&H J-class Challenger #1510 blew up accending the westbound Richmondville 1.7% (ruling) grade in 1941 I believe it was. The boiler was blown ahead nearly 300 feet and the train collided with it I remember reading.There is a couple of pictures of it in Shaughnessy's great book on the D&H
I'm sure an ICC report exisit on this accident as well.
  by Eliphaz
 
how do you test a low water alarm?
the low water alarm is typically a float in a water column. it has a body drain, just like a sight glass. to test it, open the drain so the water is diverted out of the body of the column. the float drops and operates the alarm.
  by GSC
 
That was a great report. Thanks for posting it.

We didn't run anything quite as big as the Allegheny, but we were taught that if the water went low, dump the fire and if necessary, open the safeties. Lower the pressure - safely - and cool the boiler - safely. Any boiler can cause major destruction if it blows up. Look at the Gettysburg RR a few years ago, and that steam tractor at that show in Ohio when the boilers went.

When water goes low and the crown sheet is exposed, the temperature of that steel increases quickly and begins to warp, causing thin spots and twisting and fractures and it tears open, allowing all that pressure a place to get out - and fast. Most crown sheet failures would cause a downward blast, causing, in the Allengheiny's case, the rear of boiler to lift upward away from the ground in a flipover motion, which that boiler did. Smaller boilered engines and tractors might just jump upward without breaking away from the frame, but upsetting the vehicle from where it sat. In the case of the Gettysburg RR, the firebox door happened to be open at the time of explosion and a good portion of the pressure went into the cab rather than down through the bottom of the firebox.

There were other causes of boiler failures other than crown sheets going, of course, but most explosions were caused by low water. Even though the fire in the firebox might be cooking that firebox steel at 1500 degrees or more, the water on the other side of that steel is only in the 400-500 range. A big difference.
  by johnthefireman
 
Are boilers in the USA not fitted with fusible plugs? These are common in South Africa and, I presume, elsewhere. It's a plug filled with an alloy which melts at a lower temperature than the firebox crown. Most locos have three or four of them fitted to the top of the firebox. If the water exposes the crown, the core of the plugs melts and the ensuing jets of steam extinguish the fire as well as reducing boiler pressure. It's very embarrassing for the fireman to have "dropped a plug", but it prevents a boiler explosion and is usually safe (unless the firebox door happens to be open, in which case hot coals can be blown onto the footplate, injuring the crew).
  by GSC
 
Yes, John, American locomotives are required to have fusible plugs in the crown sheet. Sorry for not mentioning that. In the case of the two explosions I mentioned, the locomotive's fusible plug, as well as the water sight glass, had been plugged up with scale. In theory, the plug fillings should melt and let water and steam into the firebox, releasing pressure and putting out the fire. At Gettysburg, they didn't. As far as the steam tractor, I do not know if there was a fusible plug in the crown sheet. These come under the federal heading of "antique and hobby boilers" and I don't know the age or inspection record, if any, of that vessel.

We had a sign in our machine shop that said it all: "Low water causes boiler explosions". Easy concept to understand.
johnthefireman wrote:Are boilers in the USA not fitted with fusible plugs? These are common in South Africa and, I presume, elsewhere. It's a plug filled with an alloy which melts at a lower temperature than the firebox crown. Most locos have three or four of them fitted to the top of the firebox. If the water exposes the crown, the core of the plugs melts and the ensuing jets of steam extinguish the fire as well as reducing boiler pressure. It's very embarrassing for the fireman to have "dropped a plug", but it prevents a boiler explosion and is usually safe (unless the firebox door happens to be open, in which case hot coals can be blown onto the footplate, injuring the crew).
  by johnthefireman
 
Thanks, GSC. I'd read somewhere that fusible plugs can get covered with scale and become inoperable. In South Africa they are inspected regularly. A couple of years ago I had the opportunity to watch an old boiler inspector refilling plugs with fresh alloy - an interesting process. I posted some photos at http://www.friendsoftherail.com/phpBB2/ ... ugs#p24142
  by Frisco1630
 
GSC. Surprised by this comment. It has certainlky not been my experience. The report on the Gettysberg failure specifically states-

Federal regulations do not require the use of fusible plugs but do require that if the plugs are used, they must be maintained. According to 49 CFR 230.14 (so at the 31 day inspection) and using the procedure set out in 230.59).

Having been involved in steam preservation before and since my move to the US 12 years ago, the absence of plugs in any of the fireboxes that I have worked in here is a very noticable difference.