Revision history [back]
 | 1 | initial version |
Gmax is the maximum gate opening in pu. Maximum gate opening is 1 pu, i.e. gmax should normally be set to 1 pu.
The turbine power, Pmech, is in pu Mbase, i.e. a different pu system. In HYGOV parameter qNL defines the turbine flow at zero turbine power, normally 0.08 - 0.1 pu flow. At steady state conditions (constant frequency) flow q has the same value as gate opening g.
The turbine gain, At and qNL, are used to convert the signal from pu flow to pu power. The following figure explains the relationship:
Maximum Pmech: Pmaxpu = Pmax(MW)/Mbase(MVA).
The turbine gain At is used to match maximum Pmech to Gmax, using this formula:
At = Pmax/Mbase/(Gmax-qNL)
Example: Pmax 95 MW, Mbase 100 MVA, qNL = 0.1:
At = 95/100/(1-0.1) = 0.95/0.9 = 1.0555556
Parameter At is the slope of the line in the figure above.
Pmin in the figure just illustrates that the minimum turbine power in HYGOV is negative at zero gate opening (zero flow).
Here, gate opening 1 pu corresponds to (1-0.1)x1.0555556 = 0.95 pu Mbase = 95 MW, i.e. dynamic limit is the same as the load flow limit.
If Rsource > 0 for the generator the machine losses makes Pmech a little higher than 0.95 pu at maximum generation in load flow. The turbine gain should then be increased a little in order to avoid initialisation problems with HYGOV model.
So, Gmax = 1 is the correct limit in PSSE. Parameter At shall be calculated for each individual hydro generator.
===========================================================
Warning: PTI manual (PAG-II, chapter 18.5.2) presents the following "representive data" for HYGOV:
Gmax = 1, qNL = 0.08, At = 1.2
Here, maximum gate opening corresponds to Pmech = (1-0.08)x1.2 = 1.104 pu. For our generator that is 110.4 MW compared to maximum power 95 MW in load flow. Hence, the dynamic model may wrongly heavily overload the turbine at underfrequency situations.
There are more examples in the manual where PTI recommends too high dynamic limits for governor models. E.g. GAST model, here PTI recommend VMAX = 1 pu, which is wrong if Pmax < Mbase.
Many PSSE users worldwide are thereby using wrong power limits in their dynamic models.
