eCathode coating adhesion›Monitoring & Alerts›Pilot coating line 2
Target experience · needs ADR-0006 sign-offToday · control is advice (A2)← Design setIllustrative values
Pilot coating line 2Project section · Monitoring & Alerts
Monitoring & Alerts
Advisory1 open incident · 1 monitor live · 2 coworkers on trigger · data fresh11:06:14 IST
Live signals
live · Line historian · OPC UA gateway (read-only)
Bands from process specification PS-L2 rev D · select a signal to see where it comes from
Incident 9 · Zone 3 anneal advisory
open since 10:45 · acknowledged · control plan waiting for approval
Anneal watch v2monitor · L0raised 10:54
AdvisoryAlert record · Incident 9
Zone 3 at 603 °C for 9 min during roll 17-04, above the 595 °C advisory band
Rule
anneal-watch/zone3-advisory revision 3 · TC-3B > 595 °C for ≥ 9 min
Evidence window
10:45–10:54 · 540 samples at 1 Hz ·
Band source
Process specification PS-L2 rev D §3.2 · normal ≤ 595 · advisory 595–605 · alarm > 605 °C
Linked
EElidyneexplanation · answered from records10:58grounded in 4 records
Why is zone 3 above its band at an unchanged setpoint?
The setpoint has not moved: zone 3 has been at since the process change. At 10:41 the web speed dropped from to , which lengthens the dwell in zone 3. For 1.3 m/min the twin predicts a strip temperature of , against at 1.5 m/min. TC-3B read at 10:54: the residual is against the 1.3 m/min prediction, inside the twin's 95 % band, and against 1.5 m/min, outside it.
The companion thermocouple TC-3A reads , within 1 °C of TC-3B, so this is not a sensor fault. Heater duty has held at 64 % since 10:41, so the oven is not overshooting; the web is spending longer in it.
Likely explanation, not a conclusionThe twin re-run at 1.3 m/min and the PT-200 test on roll 17-04 decide it.
Figure 1. Zone 3 temperature (TC-3B), 10:30–11:06, against the twin's 95 % bands at both web speeds. Shaded column: alert window 10:45–10:54. The dashed band is the twin at 1.3 m/min, confirmed by Run 1 at 11:04.
Hover the figure, or focus it and use the arrow keys, to read one sample.
Checked and set aside
Sensor fault: TC-3A reads 602 °C, within 1 °C of TC-3B, and neither shows drift or spikes over the last 6 h.
Heater runaway: zone 3 heater duty 64 %, steady since 10:41; zones 2 and 4 on their setpoints.
Elsewhere on the line: coating thickness 62 µm and dry-room humidity 1.1 % RH, both normal; nothing else moved at 10:41.
Not ruled out: adhesion on roll 17-04. The last PT-200 (0.47 N/mm) is from roll 17-03, tested 10:32; roll 17-04's test comes after the roll.
Updated 11:04 by Run 1 of “Oven twin at 1.3 m/min”: the re-run confirms this explanation, residual +2 °C. It stays an explanation until the PT-200 result on roll 17-04.
1 Historian window 10:45–10:54 · 2 Twin prediction, Investigation 12 · 3 Web speed record, 10:41 · 4 Process specification PS-L2 rev D · 5 Run 1, Oven twin at 1.3 m/min
EElidyneproposal10:59
L1execute with approval · spawns an investigation from this incident
Re-run the oven twin at 1.3 m/min
Start investigation “Oven twin at 1.3 m/min” · plan v1 · 0.6 CU · runs on the shared CPU pool
Set the oven thermal model to the observed conditions: web speed 1.3 m/min, zone 3 setpoint 585 °C, heater duty 64 %
Solve the strip temperature through zones 1–4 with its 95 % band
Compare the zone 3 prediction with the historian window 10:45–10:54
Readstwin · 4 streams · 1 window
Writesa twin revision, not adopted
Estimate0.45–0.6 CU · reserved at p90
Approved by you · 11:00 · 0.6 CU reserved
Run 1 · Oven twin at 1.3 m/minfinished 11:042 m 10 s · 0.5 CU of 0.6 reserved · shared CPU pool
Twin re-run finished: predicted at 1.3 m/min → residual , inside band
The 1.3 m/min operating point is recorded as twin revision r2, not adopted. Anneal watch v2 keeps reading the raw thermocouple; v3, in shadow, reads the twin's strip temperature.
Bring zone 3 back inside the specification and hold roll 17-04 for its peel test
Steps
One press approves the whole plan. Nothing runs until you execute it.
EElidynemonitor review11:06
Shadow · never acts
Anneal watch v3 (proposed, running in shadow since 06:00 on 21 Sep · never acts)
Compared with v2, live, on the same 5 streams over the last 24 h (11:06 on 21 Sep to 11:06 today). v3 reads the twin's strip temperature instead of the raw thermocouple.
Last 24 h
v2 live
v3 shadow
Alerts raised
6
2
Later cleared as noise
4
0
Valid alerts missed
0
0
Zone 3 advisory raised
10:54
10:51 · 3 min earlier
valid alertcleared as noise
›What changed in v3 · 3 rule changes
Zone advisory: the twin's strip temperature (from TC-3B, web speed and heater duty) above 595 °C for 6 min, instead of raw TC-3B above 595 °C for 9 min. Air overshoots at roll changes and door purges stay quiet.
Thermocouple spikes: TC-3A and TC-3B must agree within 3 °C before any zone rule fires. Single-sample glitches stay quiet.
New combined rule: web speed below 1.4 m/min for 5 min with zone 3 above 590 °C. This is what raised today's advisory 3 min earlier.
›All 6 alerts, side by side
When
What
v2
v3
Outcome
Promotion needs your approval · v3 stays at L0
Promote Anneal watch v3 to live? v2 stops raising alerts and moves to shadow for 7 days. v3 stays at L0 suggest only, and its standing approval is unchanged: may read 5 streams and the twin, ≤ 0.2 CU per firing.