Capstone: a safety device test and a door entrapment call

The capstone for this term was two pieces of work in the same week, and Lena chose them deliberately. The first was planned months ahead: Quayside car 2's five-year tests of its governor, safeties, buffers and emergency brake, with the inspector present. The second was not planned at all: a door entrapment call at Larchmont at seven in the evening on the Thursday. "One is the most controlled job we do and the other is the least," Lena told Tamsin. "You need to be good at both, and they use the same knowledge." This lesson follows both jobs from start to finish and ends with Tamsin's write-ups, which became part of her apprenticeship record.

Part 1: planning the five-year tests

Five-year tests are full-load, rated-speed tests of the devices that stop a runaway car, so they are planned in writing. Priya's plan for Quayside car 2 had five parts. First, the scope: governor tripping and switch speeds and pull-through, car safety test with rated load at rated speed, counterweight safety test, car and counterweight oil buffer tests at the speed allowed by the emergency terminal speed limiting system, and the unintended movement and ascending overspeed tests by the maker's procedure. Second, the logistics: test weights for the rated load 1,600 kg (3,500 lb), delivered and moved in the car on a Sunday morning when the building's traffic is lowest, with the other three cars in service for residents; barricades at the car 2 entrances; the inspector booked. Third, the method: the maker's procedures for each test, set up through the service tool so that the safety board's supervision is suspended only for the device under test and restored afterward. Fourth, the hazards: working on a car top near moving equipment, the weights, the high-speed tests, and the possibility of a device failing to act, each with its control in the job hazard analysis. Fifth, the return to service: every device restored, every bypass removed, the car run on automatic through every floor, and the test record signed.

# Five-year test day plan for Quayside car 2: sequence and time
steps = [("barricade, lockout, job hazard analysis briefing", 30),
         ("governor: rope off, speed test with test drive, pull-through, reseal", 75),
         ("load test weights into car", 40),
         ("car safety test: rated load, rated speed, governor tripped", 35),
         ("raise off safeties, reset switch, measure marks, dress rails", 45),
         ("car buffer test at limited speed, check return and oil", 30),
         ("unload weights", 35),
         ("counterweight safety and buffer tests, car empty", 50),
         ("unintended movement and ascending overspeed tests by maker's procedure", 40),
         ("restore all devices, end service tool tests, remove bypasses", 25),
         ("run every floor on automatic, door checks, sign the record", 30)]
t = 0
for name, minutes in steps:
    print(f"{t // 60:2d}:{t % 60:02d}  +{minutes:3d} min  {name}")
    t += minutes
print(f"total about {t / 60:.1f} hours, starting 07:00 -> finished by about {7 + t // 60}:{t % 60:02d}")

 0:00  + 30 min  barricade, lockout, job hazard analysis briefing
 0:30  + 75 min  governor: rope off, speed test with test drive, pull-through, reseal
 1:45  + 40 min  load test weights into car
 2:25  + 35 min  car safety test: rated load, rated speed, governor tripped
 3:00  + 45 min  raise off safeties, reset switch, measure marks, dress rails
 3:45  + 30 min  car buffer test at limited speed, check return and oil
 4:15  + 35 min  unload weights
 4:50  + 50 min  counterweight safety and buffer tests, car empty
 5:40  + 40 min  unintended movement and ascending overspeed tests by maker's procedure
 6:20  + 25 min  restore all devices, end service tool tests, remove bypasses
 6:45  + 30 min  run every floor on automatic, door checks, sign the record
total about 7.2 hours, starting 07:00 -> finished by about 14:15

The plan came to about seven and a half hours, and Priya booked the whole day. The order matters: the governor is tested before the safeties because the safety test uses the governor's trip, the car tests are done with the weights in, and the counterweight tests with the car empty, so the weights are moved only once.

Part 1: the test day

The governor tests came first. With the governor rope lifted off the sheave and a test drive turning the sheave, the tripping speed came in at 3.05 m/s (600 ft/min), within the code range for the car's rated speed, and the overspeed switch, adjusted in lesson 7, opened below its maximum. Three pull-through readings averaged within the maker's range. The governor was resealed.

With the rated load in the car, the car was run down from the top at rated speed and the governor tripped by hand at the governor, from the machine space, on Priya's radio call. The car stopped smoothly. The marks on the rails measured 1.22 m and 1.18 m, close together, so the platform was level within the limit, and the stopping distance was within the maker's range for the tripping speed. The car was raised off the safeties on inspection, the safety switch reset, and Tamsin dressed the rail marks with a stone. The car buffer test at the speed the terminal speed limiting system allows followed: the plunger returned in well under the time allowed and the oil level was correct. After the weights were unloaded, the counterweight safety and buffer tests were done with the car empty; the counterweight buffer's oil was found slightly low, the same finding as in lesson 10's record, and was topped up with the specified oil and a small seep at the plunger seal noted for repair.

# Quayside car 2 five-year results against their criteria
results = [
    ("governor trip speed", 3.05, 2.88, 3.18, "m/s"),
    ("governor switch speed", 2.8, 0, 2.86, "m/s"),
    ("pull-through average", 2480, 2200, 2900, "N"),
    ("car safety stopping distance", 1.22, 0.8, 1.6, "m"),
    ("platform tilt", 2.1, 0, 3.0, "%"),
    ("car buffer return time", 41, 0, 90, "s"),
    ("unintended movement stop", 260, 0, 1220, "mm"),
]
fails = 0
for name, value, lo, hi, unit in results:
    ok = lo <= value <= hi
    fails += not ok
    print(f"{name:<30} {value:>8} {unit:<7} (allowed {lo} to {hi}) {'pass' if ok else 'FAIL'}")
print("all tests passed" if fails == 0 else f"{fails} test(s) failed: car stays out of service")

governor trip speed                3.05 m/s     (allowed 2.88 to 3.18) pass
governor switch speed               2.8 m/s     (allowed 0 to 2.86) pass
pull-through average               2480 N       (allowed 2200 to 2900) pass
car safety stopping distance       1.22 m       (allowed 0.8 to 1.6) pass
platform tilt                       2.1 %       (allowed 0 to 3.0) pass
car buffer return time               41 s       (allowed 0 to 90) pass
unintended movement stop            260 mm      (allowed 0 to 1220) pass
all tests passed

The stopping-distance range and the tilt limit in the record are the maker's figures for this car and safety, taken from the test procedure, and the inspector checked them against the code table. Every result passed. The unintended movement test, run through the service tool with the doors open at floor 1, stopped the car a small fraction of the code distance from the floor. The inspector witnessed the safety and buffer tests and signed the record; the test tag in the machine space was updated.

Returning the car to service took longer than Tamsin expected. Priya ended each service tool test so that the safety board restored its supervision, checked the controller showed no bypasses active, ran the car to both terminals on inspection to prove the final limits and normal terminal devices, then put it on automatic and rode every floor, checking the doors at each landing. Only then were the barricades removed.

Part 2: the entrapment call

At 18:52 on the Thursday, the monitoring centre took an emergency call from Larchmont's car: two residents and a dog, moving a bookcase, were stopped in the car near floor 5 with the doors closed. The centre identified the car, now that its device had been fixed in lesson 11, kept talking to the passengers, confirmed nobody was hurt or unwell and called Northcrest's after-hours line. Tamsin was on call with Ray as her backup and reached the building at 19:24. She spoke to the passengers through the building's in-building communication, told them who she was and what would happen, and asked them not to try to open the doors. Then she followed Northcrest's entrapment procedure.

# Entrapment response timeline (times from the monitoring centre and Tamsin's log)
events = [("18:52", "call received, car identified, passengers reassured"),
          ("18:55", "Northcrest after-hours line called"),
          ("19:24", "mechanic on site, talks to passengers"),
          ("19:31", "car position found: 180 mm above floor 5, in unlocking zone"),
          ("19:36", "mainline locked out, hoistway door unlocked with tool, barricade up"),
          ("19:39", "car door opened by hand, passengers out over the sill with help"),
          ("19:55", "cause found: floor 5 hoistway door panel out of its sill groove")]
def minutes(t):
    h, m = map(int, t.split(":")); return h * 60 + m
start = minutes(events[0][0])
for t, what in events:
    print(f"{t}  (+{minutes(t) - start:2d} min)  {what}")
print(f"passengers out {minutes('19:39') - start} min after their call")

18:52  (+ 0 min)  call received, car identified, passengers reassured
18:55  (+ 3 min)  Northcrest after-hours line called
19:24  (+32 min)  mechanic on site, talks to passengers
19:31  (+39 min)  car position found: 180 mm above floor 5, in unlocking zone
19:36  (+44 min)  mainline locked out, hoistway door unlocked with tool, barricade up
19:39  (+47 min)  car door opened by hand, passengers out over the sill with help
19:55  (+63 min)  cause found: floor 5 hoistway door panel out of its sill groove
passengers out 47 min after their call

The car's position decided the method. The controller showed the car stopped about 180 mm above floor 5, inside the unlocking zone, with the door circuit open. Because the car was in the zone, the passengers could be released at floor 5 without moving the car. Tamsin locked out the mainline disconnect, so that the car could not move during the rescue, opened the floor 5 hoistway door with the unlocking tool from the landing, set up a barricade, and opened the car door by hand; the restrictor allowed it because the car was in the zone. The step down from the car sill to the landing was small, and she helped the residents, the dog and finally the bookcase out. Had the car been outside the zone, the procedure would have been to move it to a floor first, by the maker's rescue operation, not to open the doors between floors.

Part 2: finding and fixing the cause

With the passengers out and the car locked out, Tamsin looked for the cause. The floor 5 hoistway door's leading panel had a cracked gib, and the bottom of the panel had been pushed out of the sill groove toward the hoistway, probably by the bookcase or a cart earlier in the day. As the car arrived at floor 5 and its door began to open, the clutch picked up the displaced panel's interlock rollers out of line; the hoistway door jammed partway, the car door closed again on the controller's door-close retry, and when the controller tried to relevel with the door circuit not proving locked, it shut the car down. The car had stopped in the door zone, as it should, and the restrictor and interlock had kept the passengers from opening the doors themselves.

The repair followed lesson 3 and lesson 6. With the car moved to floor 6 on inspection and the floor 5 entrance barricaded, Tamsin and Ray, who had arrived by then, replaced both gibs on the panel with listed fire-resistant gibs, refitted the panel in the sill, checked the hangers and up-thrust rollers, checked the panel gaps, checked the interlock engagement and contact sequence and the clutch vane clearance at floor 5, and checked the floor 5 closer. Then they ran the tests: door closing force and time at floor 5 against the data plate, the door from every position by hand with the car away, the interlock opened by the unlocking tool to prove the car would not run, and twenty door cycles on automatic. The car returned to service at 21:40.

# Entrapment report checklist: is anything missing before it is filed?
report = {"time of call and of release": True, "car and landing": True, "passengers and condition": True,
          "car position and rescue method": True, "cause found": True, "parts replaced": True,
          "tests after repair": True, "reported to the regulator if required": True,
          "follow-up: cart damage and other landings checked": False}
missing = [k for k, v in report.items() if not v]
print("report complete" if not missing else "complete before filing: " + "; ".join(missing))

complete before filing: follow-up: cart damage and other landings checked

Tamsin's report was missing one item: she had not yet checked the other landings for similar damage. The next morning she and Priya surveyed every Larchmont landing's gibs and sills and found a second cracked gib at floor 3, which was replaced. Lena added a note to the building's file and wrote to the building manager suggesting a rule for moving furniture, with door protection pads, booked through the manager. Technical Safety BC reporting rules depend on whether anyone is injured and on the type of incident; Lena checked them and filed what was required.

Tamsin's write-up

Tamsin's apprenticeship write-up compared the two jobs. The five-year test proved, under control, that the devices of lessons 7 to 10 would stop the car in the failures they are designed for. The entrapment showed the door devices of lessons 3 to 5 doing their job in a real failure: the interlock and car door contact stopped the car from moving with a door not locked, the car stopped in the door zone, and the restrictor kept the passengers from opening the car door themselves, while the emergency communication of lesson 11 kept them informed. Her conclusion, which Priya underlined, was that in both jobs the work was only finished when every device was proven by test after the work was done.