60nF

ie
IDEAL
TRANSFORMER
p
13.
R
M
A
R
Y
MEASURED iRCFl (SOLID CURVES) ANO
CALCULATED (DASHED CURVES) |RCF|
FOR PT 72 WITH
Equivalent Circuit for the PT72 CpS-60nF
CS = OnF

Conclusions
FREQUENCY HERTZ)
1. At 60 Hz, the two PT's tested yielded less than 3% error for secondary
voltages from less than 0.1% to approximately 125% of rating.
2. The RCF is measured for each PT with burdens of 102, 103, and 106 3. Equivalent circuits are developed for the potential transformers.
ohms. For burdens of 103 ohms or more, the RCF is 1.00 ± .03 over RCF's calculated from these equivalent circuits are in excellent
the whole frequency range of 60 to 10 kHz. With a 102 ohm burden, agreement with those measured over a range of frequencies from
the RCF is 100 ± .03 from 60 to 6 kHz for the 7200:120 volt PT 60 to 30,000 Hz for voltages from 1% to 10% of rating.
tested and from 60 to 1 kHz for the 4800:120 volt PT tested.

March 1981, page 1013 Maximum acceleration at shaft positions ftp, /p//p1, lpMlp2, and Ip2l
Ip3fare 1.71,1.12,0.62 and 0.82 p.u./s2 respectively. Maximum average
acceleration measured over 4 revolutions (0.08 s) at shaft location/pi/
Shaft Torsional Phenomena in Ip2 illustrated by curve A (¡ii) is 0.16 p.u./s2. It will be seen thatis almost
Hz perturbation is acceleration due to shaft torsional effects
the 31

Governing Large Turbine- eliminated by locating the speed sensor at the lpMlp2 shaft position
and averaging acceleration over 4 revolutions.
Generators with Non-Linear Steam Figure B shows acceleration characteristics corresponding to full-load
rejection.
Valve Stroking Constraints It will be seen that acceleration at shaft location IpMlpl averaged over
4 revolutions gives rise to a smooth acceleration characteristic with
Hammons, Senior Member
T. J. negligible 31 Hz content.
It is concluded that:
McMaster Univ., Hamilton Ontario, Canada 1. Shaft flexibility can exert a significant influence on governing large
The paper examines shaft torsional phenomena in governing large
turbine-generators particularly when different valve stroking rates
are employed.
synchronous-generators following severe system disturbances and load 2. The effect can be minimised by careful location of the speed sensor
rejection. (or sensors) along the turbine shaft, and by averaging speed or ac¬
It is shown that shaft flexibility can exert a significant influence on celeration signals to filter out the higher frequency oscillations of
governing large synchronous-machines following grid network faults, the shaft.
and that the location of the speed sensor is a critical factor. The effect 3. Detailed analyses performed on each machine in conjunction with
can be minimized by averaging speed or acceleration signals over a pre¬ its network to ascertain whether under practical operating condi¬
set time to filter out the unwanted higher frequency shaft torsional tions system faults interacting with shaft torsional oscillations could
harmonics. cause undesirable governor response is recommended for all large
Figure A illustrates shaft acceleration at various shaft positions fol¬ turbine-generators. The performance of different machines in similar
lowing a 3-phase, 160 ms fault at the HV transformer busbar of a 1300 networks employing similar governors under identical fault condi¬
MW turbine-generator. tions differs significantly.
Oii) (ii) (i)
0.16 1.2
0.12 0.8
0.08
0.4
0.04
0.00 0.0
-0.C4
-0.4h
.0.08
-0.12 -0.8
-0.16! 0.4
time, s
Figure A. Shaft acceleration following a 3-phase, 160 ms fault at HV Short-circuit ofduration 0.16 s applied at H V transformer busbar. Speed
transf. busbar.1300 MW, 3000 rpm Machine, (i) hp turbine, p.u./s2 sensor located at lol /Ip2 turbine shaft,
(ii) ip/lp1 shaft, p.u./s2 (iii) ip1/lp2 shaft averaged over 0.08 s, p.u./s2
20
 CtiO (ii) (i) ovo
0.12 0.6
0.1C 0.4 0.5
coe 0.3 0.4
O.OG
0.04
0.2
0.1
0.3
0.2
i. ii
0 02 0.1
0.00 0.0 0.0
-0.1 -0.1
-0.2
-0.2 -0.3 load
-0.3 -0.4 rejected
0. 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
times
Figure B. Shaft acceleration following load rejection at full load.1300 close at 4.5 p.u./s when acceleration averaged over 0.08 s exceeds 0.07
MW, 3000 rpm Machine, (i) hp shaft, p.u./s2 (ii) ip/lpl shaft, p.u./s2 p.u./s2 and open at 0.225 p.u. when less than this value. Speed sensor
(iii) Ip1//p2 shaft averaged over 0.08 s, p.u./s2 Both hp and ip valves located at Ip 1/lp2 turbine shaft.

March 1981, page 1049 In the analysis, a generator is approximated by a simplified model.
The inner surface of the magnetic container or environmental shield is a
cylindrical surface with plane end plates. The permeability and the re¬
A TheoreticalAnalysis of Finite- sistivity of this container are assumed to be infinite. The straight(and
end-turn portions of the physical windings are developed on cylindrical
Length Electromagnetic Shields surfaces. The magnetic diffusion through a thick cylindrical shield is
analyzed by replacing the thick shield by a set of concentric thin sub-
of Superconducting shells. These shells are physically isolated but are inductively coupled.
It is assumed that a shell is thin compared to the mean radius of the
Turbine Generators shield so that the thin shell can be replaced by a surface current sheet.
An important feature of the present analysis is the treatment of
shield lengths. The analysis of an earlier paper is limited in that all
Martin Ashkin solid shields are assumed to have the same length which extends to the
Westinghouse R&D Center, Pittsburgh, Pennsylvania end of the machine. In the present work we treat independent shields
of arbitrary lengths thereby increasing the inherent accuracy of the
15235
analysis of a superconducting generator.
The analysis reported is an integral part of a program funded by the The equivalent circuits and their lumped parameters which represent
Electric Power Research Institute to design and build a 300 MVA super¬ the shields are conveniently found by first considering a single shell.
conducting turbine generator (SCG). Generator system performance This subdivision is possible since the machine being considered is linear
evaluations which include stability analysis, fault studies and unbalanced and superposition holds. Circuits for a single shell are found which are
operations use results from this analysis. mutually uncoupled but inductively coupled to circuits of the other
An equivalent circuit representation of solid eddy-current shields of a shells and to the windings. These couplings or mutual inductances are
SCG is developed. Expressions for the circuits parameters are derived calculated by either pairing one shell with a second shell or with a
and evaluated for representative machines. Transient simulations using physical winding.
these parameters will be presented in a later paper. The circuits are found by solving a three-dimensional electromag¬
The SCG under development is a machine with a stationary normal netic boundary value problem for a shell in a magnetic container.
metal armature winding and a rotating superconducting field winding. This problem is formulated in terms of differential equations for the
There is no magnetic material in the windings. Attached to the rotor be¬ Fourier coefficients of the shell surface current. The equations are
tween the field and the armature are one or more electromagnetic solved in the form of an eigenvector expansion; the coefficients in this
shields essential for continuous operation during transient or unbalanced expansion are the currents flowing in the eqivalent circuits.
armature conditions (see Figure 1).

March 1981, page 1059

Radio Interference with the

Operation of Static Utility
Protective Relays
R. L. Coren, Member, IEEE
Electrical Engineering Dept., Drexei University,
Philadelphia, PA 19104
Introduction
WarmEddy Current Shield Controls used in electric power distribution systems are generally
LCold Eddy Current Shield tolerant of the high level of interference signals found in that environ¬
ment. Recently, however, increased power demands have required a
Figure 1. Cross-sectional schematic of superconducting turbine generator. new class of solid state controls, which are more susceptible to Radio

21