1

Voltage Control Strategies for Distribution Systems with High Penetration of Photovoltaics

Reinaldo Tonkoski, Ph.D.Associate Professor

Electrical Engineering and Computer Science Dept.South Dakota State University

Brookings, SD, USAtonkoski@ieee.org

8/22/2018

2

Outline

Background Basic Concepts Classification of Inverter Control Strategies Case Studies Comparison of Inverter Control Strategies Current Trends Conclusions

3

Rising Trend of Small-scale Photovoltaic Generation

Distributed PV Generation in US (2014-2016) Estimated small-scale PV Generation

Source: U.S. Energy Information Administration, Electric Power Monthly

Small-scale PV generation accounted for 37% of annual generation from solar in U.S.

Residential installations increasingDecreasing costTrend of clean energy source

4

Inverter Control Strategies for Voltage Support

• Unidirectional flow of power

• Protection and control available for unidirectional power flow

• High PV penetration and low load

• Reverse power flow causes voltage rise

• PV power curtailed• Limits PV installation

capacity

Distance from feeder

Vol

tage

Lower Limit

Upper Limit

Legacy devicesLTCs, voltage regulators, capacitor banksFails under bidirectional power flow

Control for voltage regulation at both distribution and DG connection points required

Inverter based voltage control strategies needed to maintain power quality under high PV penetrations

Δ𝑉𝑉 ≈𝑃𝑃𝑃𝑃 + QX

𝑉𝑉

5

The Beginning!

X 9 X 9 X 9 X 9

X 9 X 9 X 9

A

C

X 9 X 9 X 9

X 9 X 9 X 9

X 9

X 9

X 9 X 9 X 9

X 9 X 9 X 9

X 9

X 9

B

20 m

20 m

20 m

20 m

20 m

20 m

20 m

20 m

20 m

2 km

20 m

20 m

20 m

Substation

250 m 250 m 250 m 250 m

1 km

1 km

A B

C

1 km

0 AWG ASC

336

kcm

il AS

C

20 m

3 MVA PF 0.90

3 MVA PF 0.90

2 MVA PF 0.95

2 MVA PF 0.95

2 MVA PF 0.95

Sub-Network A

0 AWG ASC

1/0 AWG, Aluminum, XLPE

4/0 AWG, Aluminum, XLPE

75 kVA14.4 kV/120 V/240 V

94 MVA120 kV/25 kV

0.95

0.96

0.97

0.98

0.99

1

1.01

1.02

1.03

1.04

1.05

1.06

1.07

1.08

-75 -50 -25 0 25 50 75

Volta

ge [p

u]

PMPPT-Load[kW]

Transformer

H 1/2

H 3/4

H 5/6

H 7/8

H 9/10

H 11/12

APC

- R. Tonkoski, D. Turcotte and T. H. M. EL-Fouly, "Impact of High PV Penetration on Voltage Profiles in Residential Neighborhoods," in IEEE Transactions on Sustainable Energy, vol. 3, no. 3, 2012.- R. Tonkoski, L. A. C. Lopes and T. H. M. El-Fouly, "Coordinated Active Power Curtailment of Grid Connected PV Inverters for Overvoltage Prevention," in IEEE Transactions on Sustainable Energy, vol. 2, no. 2, 2011.

6

Inverter Control Technologies for Voltage Control

First Generation Second Generation

Third Generation

7

Case Studies Around the World

SMUD & NREL

• PRECISE• In-Line Power

Regulators

HECO & NREL• VROS Project• First utility to

activate volt-VAR system-wide

Dettighofen Grid• 821.3 kWp low-voltage grid• 4.5% voltage rise due to reverse power flow• APC and RPC reduced voltage rise by 3%• Peak shifting with battery storage• 0.017 USD/kWh vs 3.58 USD/kWh

Taiwan Power Company • 3750 kWp solar farm field testing• Variable power factor and APC• 6519 kWh to 269 kWh curtailment

reduction

NEDO• Power conditioning subsystem (PCS)• 80% PV penetration• 553 residential PV systems• Total of 2.1 MW• Battery use to minimize output power

loss

8

Active Power Curtailment (APC) Based on Droop

Linear Droop* Quadratic Droop**

1.042 1.044 1.046 1.048 1.05 1.052 1.054 1.056 1.058

Voltage at the point of connection [pu]

0

0.2

0.4

0.6

0.8

1

Pow

er b

eing

cur

taile

d [p

u]

ΔV

ΔP

m=ΔP/ΔV

1.042 1.044 1.046 1.048 1.05 1.052 1.054 1.056 1.058

Voltage at the point of connection [pu]

0

0.2

0.4

0.6

0.8

1

Pow

er c

urta

iled

[pu]

y = 3906.2(x-1.042)2

𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 − 𝑚𝑚 𝑉𝑉 − 𝑉𝑉𝑐𝑐𝑐𝑐𝑖𝑖 2𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 − 𝑚𝑚(𝑉𝑉 − 𝑉𝑉𝑐𝑐𝑐𝑐𝑖𝑖)

𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖: Power injected by PV inverter𝑃𝑃𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀: Maximum power available from PV array𝑉𝑉𝑐𝑐𝑐𝑐𝑖𝑖 : Voltage above which controller comes into action𝑚𝑚: Slope factor or droop constant𝑉𝑉: Local Voltage at the point of connection

* R. Tonkoski, L. A. C. Lopes, and T. H. M. El-Fouly, “Coordinated active power curtailment of grid connected PV inverters for overvoltage prevention,” IEEE Trans. Sustain. Energy, vol. 2, no. 2, pp. 139–147, Apr. 2011. ** M. Maharjan, “Voltage regulation of low voltage distribution network”, MS thesis, Dept. Elect. Eng. Comp. Sci., South Dakota State University, South Dakota, USA, 2017

Curtailing the active power is one of the solution to prevent the overvoltage in Low Voltage (LV) network

Δ𝑉𝑉 ≈𝑃𝑃𝑃𝑃𝑉𝑉

9

Approaches to Reactive Power Control (RPC)

S. Pukhrem, M. Basu, M. F. Conlon, and K. Sunderland, “Enhanced network voltage management techniques under the proliferation of rooftop solar PV installation in low-voltage distribution network,” IEEE Journal of Emerging and Selected Topics in Power Electronics, vol. 5, no. 2, pp. 681–694, Jun. 2017.

10

Coordinated Active Reactive Power Support

10F. Olivier, P. Aristidou, D. Ernst, and T. V. Cutsem, “Active management of low-voltage networks for mitigating overvoltages due to photo-voltaic units,” IEEE Transactions on Smart Grid, vol. 7, no. 2, pp. 926–936, Mar. 2016.

normal operation

local reactive power support

neighborhood reactive power support

neighborhood active and reactive power support

QP

MPP

Q

PMPP

QP

Not at MPP

V < Vkick V ≥ Vkick

V < VkickQ = 0

tstate ≥ Δt

V < VkickPcurt = 0

tstate ≥ Δt

V ≥ VkickQi = QiMAX

V ≥ VkickQ = QMAX

11

PV Power Computation Using Solar Irradiance Data

One-hour resolution irradiance data from [18] converted into minutes

PV power computed using irradiance𝑃𝑃 = 𝜂𝜂 × 𝐼𝐼 × 𝐴𝐴 𝜂𝜂 : efficiency (16.7%), 𝐴𝐴 : area

(50,2605 𝑚𝑚2 ) [19], 𝐼𝐼 : irradiance 𝑊𝑊𝑚𝑚2

11* Clean power research, “Solar anywhere,” 2015. [Online]. Available: http://www.solaranywhere.com** https://www.solarelectricsupply.com/solarworld-8-4kw-sunmodule-plus-sw-280-mono-fronius-solar-system

Irradiance for first day of January

Irradiance for whole year

Substation

Transformer75 kVA

14.4 kV / 240 V

H1

20 m

H3

H4

20 m

H5

H6

20 m

H7

20 m

H9

H10

20 m

H11

H12

12

Comparison Year long voltage

profileOVP, LDAPC and

QDAPC maintain the voltage below limit

ARPM cannot fully eliminate overvoltage

Energy loss for a yearCurtailment energy

loss is lowest in ARPM because of the use of reactive power,

However this increases loss in the feeder and transformer

12

R. Mahat, K. Duwadi, F. B. Dos Reis, R. Fourney, R. Tonkoski, and T. M. Hansen, “A Long-Term Techno-Economic Analysis of PV Inverter Controllers for Preventing Overvoltage in Low-Voltage Grids.” [under preparation]

No overvoltageprotection

Overvoltageprotection

Lineardroop APC

Quadraticdroop APC

Active reactivePower

management

13

Centralized Control Achieve best coordination

between available controllable sources DG inverters, tap

settings, capacitors banks etc.

e.g. optimal dispatch of PV inverters

Need for solving non-convex, non-linear optimizationObjective is to reduce

curtailment, loss and maximize PV injection

Heavily dependent on communication network

S. Paudyal, C. A. Canizares and K. Bhattacharya, "Optimal Operation of Distribution Feeders in Smart Grids," in IEEE Transactions on Industrial Electronics, vol. 58, no. 10, pp. 4495-4503, Oct. 2011.

14

Distributed Control

No central controller Implemented with

computations local to nodeAugmented through limited

information from nearby nodes

Cooperation among nodes possible For example – power

loss minimization An example*: Information collected from

few inverters operating as agents

Information shared on a common cyber layer

Feedback signal approach to find optimal reactive power requirement

* S. Bolognani, R. Carli, G. Cavraro and S. Zampieri, "Distributed Reactive Power Feedback Control for Voltage Regulation and Loss Minimization," in IEEE Transactions on Automatic Control, vol. 60, no. 4, pp. 966-981, April 2015.

15

Comparison of Local and Centralized Control Method

Local active power control

Centralcontrol

Centralcontrol

Local active

reactivepower control

Controlling reactive as well as active power reduces curtailment significantly compared to active power control onlyBoth in local and central control

Central control benefits from the communicationsminimum curtailment possible

K. Duwadi, F. B. Dos Reis, R. Fourney, R. Tonkoski, and T. M. Hansen, “Optimal Inverter dispatch (OID) and time constrained OID in low voltage distribution network by leveraging linearized approximate power flow” [under preparation]

16

Neural Network Based Online Droop Adjustment

A supplementary controller to set droop in APC method

Reinforcement learning based approachA reward signal is designed Restricts voltage crossing the

critical voltage limit with minimum curtailment possible

Objective function was designed to minimize curtailment in each houseLimit voltage with critical limit Inject maximum available

power

M. Maharjan, R. Tonkoski, et al., "Adaptive droop-based active power curtailment method for overvoltage prevention in low voltage distribution network," 2017 IEEE International Conference on Electro Information Technology (EIT), Lincoln, NE, 2017, pp. 1-6.

17

Reduced Curtailment with Online Droop Adjustment

Reduced Peak Curtailment

Voltage within Critical Limit

Linear APC Adaptive APC

Energy Loss = 22.72 kWh

Energy Loss = 18.77 kWh

18

Current Trends

Voltage issues with connection of large loads in a distribution system with high PV penetration

Distributed optimization*Benefits of reduce curtailment Avoiding the dependences of centralized approachedCoordination of controllers OLTC

Multiple Agents**Attending the privacy of costumersEquilibrium of the game bids

Tackling the challenges Privacy of customers Fairness of participants Markets in the distribution system

* E. Dall’Anese and A. Simonetto, “Optimal Power Flow Pursuit,” IEEE Transactions on Smart Grid, vol. 9, no. 2, pp. 942–952, Mar. 2018.** X. Wang et al., “Optimal voltage regulation for distribution networks with multi-microgrids,” Applied Energy, vol. 210, pp. 1027–1036, Jan. 2018.

19

ConclusionsVoltage Droops are still the most popular method for

active and reactive power support for overvoltage prevention

Coordination of active and reactive power can lead to reduction on curtailment of PV

Communication infrastructure is required to optimize the performance of local active and reactive power support

Supplementary controllers can help reduce these requirements and get closer to optimal solution.

20

Thank you and Acknowledgments Drs. Timothy M. Hansen and Robert Fourney Graduate students: André Luna, Fernando Bereta dos

Reis, Kapil Duwadi and Ujjwol Tamrakar Grants:South Dakota Board of RegentsNSF ECCS-1608722NSF Major Research Instrumentation MRI-1726964

21

Thank You!