Enabling High Penetrations of Solar PV for

Southern California Edison(A Project of California Solar Initiative RD&D Solicitation #4)

Seattle, WA11/4/2016

Kevin Schneider, Ph.D., P.E.

Overview

Part 1: Project Overview

Part 2: Evaluating Circuit Native PV Limit

Part 3: Mitigating PV Violations

Part 4: Concluding Comments

2

Part 1: Project Overview

3

CSI RD&D Program Key Principals

Improve the economics of solar

technologies by reducing

technology costs and

increasing system performance

Focus on issues that directly

benefit California, and that may

not be funded by others

Fill knowledge gaps to enable

successful, wide-scale

deployment of solar distributed

generation technologies

4

Project Goals

Better understanding current grid limits for solar

penetration (native limits).

Develop technology strategies for California feeders

to obtain 100% PV penetration.

Create a cloud-based tool to study and analyze solar

PV feeder limits.

These three goals should help reduce the time and

cost required to integrate high penetration of PV on

numerous feeders.

5

100%

Get Solar PV

penetration in

California to

Know

current

system

limits

Determine

Path forward

Study Process

6

ClusterDetermine

Representative Circuits (RC)

ModelCreate RC GridLAB-D

models

Native LimitsDetermine using PV adoption study and

Monte Carlo

Mitigation TechnologiesCreate Upgrade

paths & Cost estimates

30 representative circuits

were determined using

K-Means clustering. (15

of the most

representative were used

in this study)

Circuits modeled in

GridLAB-D, with

behind the meter

loads.

Models calibrated

against SCE customer

usage data.

PV adoption

models leveraged

to determine

Native limits based

on 10 operational

constraints.

Traditional and non-

traditional

mitigation strategies

developed for

circuit upgrades to

achieve 100% PV

penetration.

Project Partners

Provide distribution model, interconnection process, validation of results, and demonstration of field interconnection

Provide GridUnity software to analyze impacts, communicate to stakeholders, and manage interconnection processProvide GridUnity software to analyze impacts, communicate to stakeholders, and manage interconnection process

Determine native Solar PV penetration levels for representative feeders and identify cost-effective mitigation strategies for higher levels of Solar PV

Southern California Edison

Project SponsorsCalifornia Public Utilities Commission, California Solar Initiative, Itron

For more information, including project reports, see:

http://www.calsolarresearch.ca.gov/

Part 2: Evaluating Circuit Native PV Limit

8

SCE Defined Operational Limits (For this project)

9

Table 4.1 - Circuit Operational Limits and Thresholds For determining Native Limits

Violation # Violation Violation Description

1 Thermal OverloadsLimit: Exceeding any device thermal limit, 100% rating (200% for secondary service

transformers)

2 High Instant Voltage Limit: Any instantaneous voltage over 1.10 p.u. at any point in the system.

3 5 min ANSI Violation Limit: ANSI C84.1: 0.95>V>1.05 p.u. for 5 minutes at >10% of meters in the system.

4 Moderate Reverse PowerWarning: Any reverse power that exceeds 50% of the minimum trip setting of the

substation breaker or a recloser. (Requires analysis of protection coordination)

5 High Reverse PowerLimit: Any reverse power that exceeds 75% of the minimum trip setting of the substation

breaker or a recloser.

6 Voltage Flicker

Limit: any voltage change at a PV point of common coupling that is greater than 5%

between two one-minute simulation time-steps. (Adapted from the Voltage fluctuation

design limits, May 1994)

7Voltage Drop/Rise on

Secondary

Limit: 3V drop or 5V rise across the secondary distribution system (Defined as the high

side of the service transformer to the customer meter)

8 Low Average PF Warning: Average circuit power factor <0.85 (Measured at substation)

9Circuit Plan Loading

Limit

Warning: Nameplate solar exceeds 10MVA for a 12 kV circuit, 13 MVA for a 16 kV

circuit, or 32 MVA for a 33 kV circuit.

10High Short Circuit

Contribution

Warning: Total short circuit contribution from downstream generation not to exceed

87.5% of substation circuit breaker rating

Determining the Native PV Limit

For effective analysis of solar issues a simple

power simulation is not adequate.

Time-series simulations: one-minute time-series

simulations were conducted for one-week periods,

one for each of four seasons.

PV adoption model: For each scenario, PV

deployment was determined by past energy usage

and a socio-economic model. Uniform

distributions are not appropriate for distributed PV

scenarios.

At each penetration level between 5% and 100%,

in 5% increments, 50 scenarios are examined.

Each scenario requires 4 one-week time-series

simulations with one-minute time-steps.

The results from the 4,000 time-series

simulations need to be represented in an easily

accessible visualization.

10

Building Type (BT) Install Size (kW) Probability of Installation (2015)

BT0 – General Miscellaneous ?? ??BT1 – Agricultural 194 0.02859BT2 – Assembly 91 0.02823BT3 – Education: Primary 149 0.09471BT4 – Education: Secondary 363 0.04289BT6 – Education: Community College 232 0.01215BT7 – Education: University 118 0.00357BT8 – Grocery 47 0.00107BT9 – Food Store 18 0.00250BT10 – Hospital 318 0.00572BT11 – Nursing Home 45 0.00715BT12 – Clinic 28 0.03788BT13 – Hotel 101 0.00643BT14 – Guest Rooms 8 0.00071BT15 – Motel 23 0.00179BT17 – Manufacturing – Light Industrial 141 0.03860BT18 – Industrial 255 0.00536BT19 – Miscellaneous Commercial 131 0.06969BT20 – Office: Large 111 0.03788BT21 – Office: Small 30 0.42102BT23 – Multifamily Residential 16 0.01072BT25 – Restaurant: Fast Food 68 0.00107BT26 – Restaurant: Sit Down 19 0.00465BT27 – Retail: Multistory Large 331 0.03002BT28 – Retail: Single-Story Large 205 0.04753BT29 – Retail: Small 39 0.01465BT31 – Storage: Unconditioned 267 0.00036BT32 – Transportation, Communication, Utilities

1850.04325

BT33 – Warehouse: Refrigerated 607 0.00179

Native Limit Curve (Circuit 11)

11

Native Limit Curve (Circuit 19)

12

Part 3: Mitigating PV Violations

13

Traditional Upgrades

Shunt capacitors

Remove existing units

Adjust existing units

Install fixed units

Install var-controlled units

Voltage Regulators

Install substation units regulating

output voltage

Reconductor/upgrade

Primary conductor

Service transformers

Secondary drops

14

Solar PV Inverter Upgrades

It is assumed that all “new” solar

above the native limit is equipped

with these controls, but there is no

retrofitting.

The controls are based on the Rule 21

Volt-var plot shown to the right.

The set points for the Rule 21 control

are fixed and do not change over time.

The second of two journal papers

being written is examining the

possibility of adjusting the default

Rule 21 values so that inverters do not

have to be over-sized.

15

Battery Storage Inverter Upgrades

Substation Distributed

This is assumed to be a utility-owned asset

Controls the reactive power output of the

inverter

The control goal is to maintain the power factor

at the substation within a desired range

Control Variables:

pf_reg INCLUDED_ALT;

pf_reg_high -0.98;

pf_reg_low 0.99;

pf_target 0.98;

pf_reg_activate_lockout_time 60s;

pf_max_capacitive_Q 1000 kVAr;

pf_max_inductive_Q 1000 kVAr;

This is assumed to be a customer-owned

asset

Controls the active power output of the

inverter

The control goal is to minimize the peak

load of an individual commercial/industrial

customer

Control Variables:

charge_on_threshold -55000

charge_off_threshold -50000;

discharge_off_threshold -65000;

discharge_on_threshold -40000;

16

Mitigation Example (Circuit 11)

17

at X% PV Limiting Violations Traditional Mitigation: Added two new substation capacitors

One 600 kvar (Fixed)

One 600 kvar (VAR controlled) Reduced the size of one existing downstream capacitor (600

kvar to 300 kvar)

Path 1 Central energy storage unit in 15%

Path 2: 11 decentralized storage units in peak shaving control

Six Large Units, 250 kW/1,000 kWh

{Charge on=-55 kW Charge off=-50 kW Discharge on=500 kW Discharge off=300kW}

Five small units, 100 kW/ 50 kWh

{Charge on=-0.5 kW Charge off=0 kW Discharge on=5 kW Discharge off=0kW}

Cir

cuit

#11 The non-traditional mitigation upgrade path to address these violations:

15%

15% Low Average PF

Target pf 0.98, +/- 1050 kvar

Mitigation Example (Circuit 19)

18

at X% PV Limiting Violations Traditional Mitigation:

5% Voltage Flicker Added two substation Capacitors

15% Low Average PF One 150 kVAR (Fixed)

One 150 Kkvar (VAR controlled)

45% 5 min ANSI Violation Added one substaion regulator controlling output voltage

to 2,380V

0% Fixed power factor control with 0.95 leading

The non-traditional mitigation upgrade path to address these violations:

Cir

cuit

#19

Part 4: Concluding Comments

All 15 circuits can support 100% penetration of PV once the proper

mitigation strategies have been applied.

Nearly 50% of SCE circuits can host less than 50% PV, where approx. 40%

can host less than 25% PV

Determining how to achieve 100% penetration on legacy circuits can be

challenging, with a mitigation leading to new violations. (domino effect)

The most common violations experienced were power factor and voltage

based.

Proper sizing of secondary drops when new solar is installed is essential.

19

Concluding Comments (cont.)

Controlling circuit voltage and circuit power factor simultaneously with

capacitors is not practical at high penetrations of PV.

Energy storage is a technically viable solution for power factor, but may not

be cost effective unless it is part of a larger multi-objective control strategy.

Inverter-based Volt-VAR is not able to address low lagging power factor and

high voltages at the same time. However, Volt-VAR combined with other

traditional upgrades can be highly effective.

20

21