Ako je to s dlhodobou presnostou frekvencie v rozvodnej sieti

[Jan Waclawek]

Tak som nasiel odpoved na moju povodnu otazku

> napriklad bol ten budik v tych -15 sekundach cely tyzden, potom za jeden den dosiel na nulu a dalsi bol uz tych +15 sekund. [atd]
http://list.hw.cz/pipermail/hw-list/2017-January/496455.html

a to v dokumentoch, na ktore dal Dano link v
http://list.hw.cz/pipermail/hw-list/2017-January/496484.html (este raz
dakujem, naozaj je to rozsiahle ale aj poucne citanie, a to nielen z
technickej stranky).

Konkretne co sa tyka toho presneho casu, tak relevantne su dokumenty P1 a
A1, v oboch kapitola D. Time Control.

Takze (v poradi aby to bolo pochopitelne z pohladu mojej otazky):

D-S3. Time Deviation Calculation. The  TIME DEVIATION between  SYNCHRONOUS
TIME and UCT has to be calculated for 8 a.m. each day by the time monitor. 

D-D1. Tolerated Range of Discrepancy. A discrepancy between  SYNCHRONOUS
TIME and UTC is tolerated within a range of ±20 seconds (without need for
time control actions).

D-D2. Target Range of Discrepancy. The discrepancy between  SYNCHRONOUS
TIME and UTC is within a range of ±30 seconds under normal conditions in
case of trouble-free operation of the interconnected network.

D-S4. Time Correction Frequency Offset. If the  TIME DEVIATION is within
D-D1 (t.j. +-20s), the  FREQUENCY OFFSET for time correction has to be set
to zero. 
If the deviation is out of D-D1 [...] the  FREQUENCY OFFSET has to be set
to (+- podla zmyslu odchylky) 10 mHz. 

D-S6. Time Correction Notice. The information for the time correction has
to be forwarded towards all  CONTROL AREAS /  BLOCKS of the  SYNCHRONOUS
AREA every day at 10 a.m. by the time monitor.

T.j.
- rozdiel casu medzi "sietovym" a UTC sa vypocitava raz za den
- ak je rozdiel mensi ako 20s, tak nic
- ak je vacsi, tak sa povie vsetkym elektrarnam ze cielova frekvencia je
nie 50Hz ale 50.01Hz alebo 49.99Hz, podla smeru odchylky; a na tuto
cielovu frekvenciu sa potom ide cely nasledujuci den 24 hodin 

Riesia sa tam potom este aj vacsie odchylky ale to je zrejme vynimocne.

Samozrejme sa tych 50Hz/50.01Hz/49.99Hz nepodari drzat cely den bez zmeny,
to riesia regulacne mechanizmy kvoli premenlivemu odberu rozoberane inde v
tych normach, takze sa potom dalsi den robi dalsia korekcia atd.

Tych 10mHz urobi asi 17 sekund za den.

Ergo, moje pozorovanie tych pomerne dlhodobych -15s bolo viacmenej spravne
(a tak trocha to potvrdzuje aj pomerne dobre iduce "referencne" hodiny
riadene DCF) - je to rozmedzie ked sa kvoli casu nerobi regulacny zasah; a
napokon aj to dotiahnutie do okolia nuly a potom do plusu mohlo byt
nahodne a len nasledok regulacia kvoli vykonu.

wek