SAFETY-129 round 3: the real reason for the 16 s limit, a slow-start check, the row's tail

Co-Authored-By: Claude <noreply@anthropic.com>

goodwe_controller/DECISIONS/SAFETY-128-late-readings.md

@@ -33,9 +33,10 @@ inverter limit in its last seconds (OPT-34 rig F-2).
      16 new values in 60 s, so only a meter of about 3.5 s or faster: at 5 s
      it gave false delays. Late over 8 s. At a slower meter (a 5 s P1) the
      same fit runs over 120 s and calls only a delay over 16 s late, on 2 fits
-     in a row, and never ends a verdict (SAFETY-129): a 10-12 s slow battery
-     reads up to 15 s there, so 10 s late is not told from it, and it does not
-     swing at 5 s.
+     in a row, and never ends a verdict (SAFETY-129). Alone it reads 35-90 s on healthy
+     runs, a few times in a row; only the slow-start and battery-fit check
+     below keeps those from a verdict. 10 s late is not found: it cannot be
+     told from a 10 s slow battery, and it does not swing at 5 s.
    - A meter fit says late only with evidence from the batteries' side too:
      a request more than 100 W from the battery reading took 3 s or more to
      move a reading towards it (`SLOW_START_S`; on time they start within
@@ -71,7 +72,9 @@ The fits run all the time; the pull request has the counts.
 - Found 36-68 s after a lag starts at 1-2 s meters; at a 5 s meter 30 s late
   is found within 100 s after it starts (88 s measured) (the loop first moves 100 W after 60 s,
   SAFETY-129) with the export above -200 W, and 10 s late is not found. Until
-  then control is as before.
+  then control is as before. In a quiet house the "readings late" row stays
+  about 2 minutes longer after the lag ends (126 s against 32 s on base); no
+  harm measured, and it clears as early as on base when the load changes.
 - A battery that answers slowly is read as late readings and held: a dead
   time of 10 s or more (8 s reads 7-8; 10 s reads 10 at 1-2 s meters, 9 at
   5 s), or a 15-20 s lag (review round 2). The GoodWe measured settles in 6 s

goodwe_controller/app/lateness.py

@@ -174,8 +174,10 @@ M_SHARE, M_BETTER = 0.4, 0.7
 # SAFETY-129: a meter slower than that (a 5 s P1) has fewer than M_ROWS values
 # in M_FIT_S, so it is fitted over M_SLOW_FIT_S instead, from the same M_ROWS
 # values. It calls a delay late only over M_SLOW_LATE_S and on ON_FITS fits in
-# a row: the slow-battery runs (10-12 s dead time, R2-1) read up to 15 s at a
-# 5 s meter, and a 10 s lag cannot be told from them, nor does it swing.
+# a row. Alone it reads 35-90 s on healthy runs, a few times in a row; what
+# keeps those from a verdict is the slow-start and battery-fit check in
+# watch(), not the limit. A 10 s lag is not found: it cannot be told from a
+# slow battery (R2-1), nor does it swing at 5 s.
 M_SLOW_FIT_S = 120.0
 M_SLOW_LATE_S = 16.0
 # A meter fit counts only after an ask that took this long to start moving a

goodwe_controller/test_safety129.py

@@ -121,6 +121,15 @@ for dead in (8.0, 10.0):
     check(f"{dead:.0f} s dead time at a 5 s meter is not read as late "
           f"(R2-1 does not get worse)", secs == 0)
 
+# At a 6 s meter an 8 s dead time reads late on the slow fit alone; only the
+# slow-start and battery-fit check keeps it from a verdict (122 s late
+# without the check, from 1 084 s).
+r, secs, first = run_late(house=owner, noise=80.0, units=AB, p1_s=6.0,
+                          dead_s=8.0, dur=1500.0)
+print(f"    8 s dead time, 6 s meter: late {secs} s")
+check("an 8 s dead time at a 6 s meter is never late (the slow fit needs an "
+      "ask slow to start and no battery fit on time)", secs == 0)
+
 # A 12 s dead time is already read as late (R2-1, SAFETY-128); the slow-meter
 # fit must not make that longer. The same run with the slow fit off is the
 # tree before this card.
@@ -219,8 +228,8 @@ check("one late slow-meter fit is no fit; two in a row are one, and "
 check("the slow fit asks for a longer window and a later delay than the "
       "fast one", lateness.M_SLOW_FIT_S > lateness.M_FIT_S
       and lateness.M_SLOW_LATE_S > lateness.M_LATE_S)
-check("the slow fit's delay is above what a 10-12 s slow battery reads "
-      "(R2-1: up to 15 s at a 5 s meter)", lateness.M_SLOW_LATE_S >= 16.0)
+check("the slow fit's late limit is above the fast one's and above a "
+      "10 s lag", lateness.M_SLOW_LATE_S >= 16.0)
 
 print()
 if fails: