IEEE Power & Energy Magazine - November/December 2019 - 38

It is extremely challenging and expensive to meet changing BPS
needs by retroactively enhancing equipment that was installed
with minimal performance requirements and capabilities.
with time delays at 59.3 and 59.5 Hz. The target megawatt
value is calculated based on system net loading at the time,
and available distribution circuits are automatically armed
every 15 min to achieve the target net load level. An arming
tolerance percentage and value are depicted as well as the
difference between the available megawatt arming and the
target megawatt value. Stages 1-2 sum to 15% and stages
1-4 sum to 40% of the system net load based on the maximum allowable load shedding for N-1 and N-1-1 unit trips,
respectively. Each stage operates with an eight-cycle relay
time plus breaker operation time. A rate-of-change-of-frequency (ROCOF) load shedding element was also included
in stages 1 and 2, with a 0.5 Hz/s set point and nine-cycle
relay time plus breaker operation time.
More than 40 substations were modified, including relay
and communications upgrades, and 78% of customer circuits
(up to 70% of peak load) are participating. The adaptive
UFLS program has performed well during N-1 and N-1-1
events at various times of day and is an effective reliability improvement. Figure 3 provides an example of a severe
N-1-1 generator-loss disturbance where stages 1-4 operated
successfully to arrest the frequency and mitigate a potential
large-scale outage.

60.290

Rooftop Solar PV Deployment
California has a state mandate that 33% of its annual energy
must come from renewable sources by 2020, with a target of
60% by 2030, established by a bill, SB100, that took effect in
September 2018. However, on some days, instantaneous output from the renewable sources already exceeds 50% of the
total generation output. A large portion of that penetration
consists of BTM D-PVs, and the rapid deployment of D-PVs
is forcing changes in transmission planning, forecasting, and
grid operations at the California Independent System Operator (CAISO).
Figure 4 provides an example of the effect that D-PVs is
having on the Pacific Gas & Electric (PG&E) net load profile, driving the need for different study times and net load
levels from the CAISO's 2018-2019 transmission-planning
process. The blue curves show the net load after subtracting
the D-PVs output, which is called the managed load. The figure clearly displays a shift in the peak net load time, which,
in many parts of the state, is shifted outside the times when
solar PV is available. D-PVs penetration levels are also causing a drastic change in minimum net load levels, which need
to be carefully studied. Sensitivity analysis is also needed
to operate through and plan for ramping periods. All these

60.24 Hz
43:33.7

Steam Unit Trip
60.00 Hz
43:09.4

60.090
59.890

One Train of Dual
Train Combine
Cycle Unit Trip
59.34 Hz
43:19.3

59.690
59.50 Hz
59.490 43:10.9
59.30 Hz
43:11.8
59.290
59.090

59.085 Hz
43:29.8
Kicker 1 and 2

59.13 Hz
43:14.2

58.890

.
p.
m

.
7:

p.
m

.
7:

p.
m

.
7:

p.
m

.
7:

p.
m

.
7:

p.
m

.
:2

p.
m
43
7:

4
:2
43
7:

p.
m

.
p.
m

.
7:

p.
m

.
7:

p.
m

.
7:

p.
m

.
7:

p.
m

.
2
:1
43
7:

p.
m

.
:1

p.
m
43
7:

:0
43
7:

58.76 Hz
Stage 1 and 2 43:23.0

58.690

figure 3. The operation of adaptive UFLS for a multiple-contingency event.
38

ieee power & energy magazine

november/december 2019

IEEE Power & Energy Magazine - November/December 2019

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - November/December 2019