IEEE Power & Energy Magazine - March/April 2018 - 75

Typical DSDR Performance

Commands

Command per Second

40
30
20
10
0
16:00 16:28

17:01 17:35 18:10
Time (hh:mm)

18:45 19:22

figure 8. The voltage regulator command rate.

(including fail-safe refresh commands) typically observed
during a DSDr activation is shown in Figure 8.

Challenges and Further
Development of ADMS
after the deployment of aDMS, DEP encountered additional challenges, an awareness of which should benefit other
utilities. Using multiple grid optimization modes (i.e., loss
optimization, DSDr, and emergency) with different operational objectives and VVO device functionality requires specific commanding for an effective transition. For instance,
attempting to change from DSDr with an already reduced
bandcenter setting and emergency mode level 2-with its
use of the regulator control's internal logic, which applies
the targeted reduction percentage (i.e., 5%) to the existing
bandcenter-could result in an unintended voltage reduction
level. the loss-of-communications fail-safe functionality
and mandatory time before being able to force the control
back to its default bandcenter (i.e., no optimization mode)
are factors that could also cause a load restrike before full
completion of the transition. DEP did attempt to define the
appropriate mode transition command "logic" during design
but didn't realize that further attention was necessary until
after attempting the commands in production.
Moreover, all the modes available through aDMS-
whether the objective is voltage reduction or power factor correction-cause a change in reactive power when activated, and
the magnitude can be significant depending on the requested
transition. after such a change, a generator may need to adjust
its output to match the resulting conditions; it also risks instability (and potential damage) if it is not allowed enough time
to do so. thus, allowable rates of change for reactive power
are defined and monitored for generators. DEP learned this
fact the hard way (through an alarm at one of its generation
plants) after completing the upgrade to two-way communication for line capacitor banks-which is more reliable and
march/april 2018

much faster-and performing a power factor target change
with its legacy Var Management System. the potential for the
same issue to occur with aDMS was recognized, and functionality was added after initial deployment that effectively
"throttles" the reactive power change rate during mode transitions. an important requirement for the enhancement to operate accurately with the voltage reduction modes is the determination of the reactive load-to-voltage dependencies for the
grid-which, of course, are utility specific.
Finally, north Carolina has seen an influx of DEr over
the past three years, led primarily by solar installations having favorable state and federal tax credits. the state now
ranks second in the country for solar energy capacity, and
DEP currently has more than 1,000 Mw of DEr connected
to its distribution grid. the impacts and challenges associated with integrating DEr are fairly well known at this
point. However, an issue that might not be as commonly recognized is the influence such integration has on the EM&V
of a peak load reduction resource. Both can be active for
certain hours during the summer season, and-depending
on the location of the captured measurements (such as at the
t/D substation) and the amount of DEr-it may be difficult
to accurately estimate the change in the grid's native load
from a planned reduction. DEP is currently analyzing data
from recent summer peak seasons in an attempt to quantify
the potential impact for DSDr.

Acknowledgment
we would like to acknowledge the contributions of Jack
Pennington, a Duke Energy Progress distribution automation and grid operations subject matter expert, who recently
retired after 29 years with the company.

For Further Reading
C. w. gellings, The Smart Grid: Enabling Energy Efficiency and Demand Response. Lilburn, ga: Fairmont, 2009.
J. Medina, n. Muller, and I. roytelman, "Demand response and distribution grid operations: Opportunities and
challenges," IEEE Trans. Smart Grid, vol. 1, no. 2, pp. 193-
198, Sept. 2010.
a. Safdarian, M. Fotuhi-Firuzabad, and M. Lehtonen,
"Benefits of demand response on operation of distribution
networks: a case study," IEEE Syst. J., vol. 10, no. 1, pp.
189-197, Mar. 2016.
J. Liu, n. Zhang, C. Kang, D. Kirschen, and Q. Xia,
"Cloud energy storage for residential and small commercial
consumers: a business case study," Appl. Energy, vol. 188,
pp. 226-236, 2017.

Biographies
Jeremy Carden is with Duke Energy, raleigh, north
Carolina.
Dragan Popovic is with the University of novi Sad, Serbia.
p&e
ieee power & energy magazine

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - March/April 2018