Sub-Module Differential Power Processing for Photovoltaic ...cemeold.ece.illinois.edu/seminars/Spring13_4_1... · PV module emulation enables controllable and repeatable PV experiment

Sub-Module Differential Power Processing forPhotovoltaic Applications

Shibin Qin Robert C.N. Pilawa-Podgurski

Department of Electrical and Computer EngineeringUniversity of Illinois at Urbana-Champaign

2013.04.01

S. Qin, R. Pilawa-Podgurski (UIUC) Sub-Module DPP for PV Applications 2013.04.01 1 / 23

Outline

1 BackgroundProblem StatementDifferential Power Processing

2 Proposed ApproachArchitectureControl

3 ExperimentExperiment SetupResult and Analysis


Background

Outline





Background Problem Statement

PV Module Mismatch Problem

PV modules are connected inseries to increase voltageSeries connected PV modulessuffers from current mismatch

Partial shadingManufacturing variabilityNon-uniform ageingcharacteristics

DC

AC


Background Problem Statement

Maximum Power Point Tracking (MPPT)Conventional Solutions

Two major conventional solutions for MPPT:DC OptimizerMicro-inverter

DC

DC

DC

DC

DC

DC

DC

AC

90W

60W

90W 97%

97%

97%

60W

90W

90W

98%233W

228W240W

Overall Efficiency: 95%(a) DC optimizer

DC

AC

DC

AC

DC

AC

90W

60W

90W 95%

95%

95%

60W

90W

90W

228W240W

Overall Efficiency: 95%(b) Micro-inverter


Background Differential Power Processing

Differential Power Processing MPPTAn Alternative Solution

The bulk power goes directly tothe central converterEach converter processes onlythe power differenceAvoid intermediate conversionof the bulk power

DC

DC

DC

DC

DC

AC

90W

90W

60W

18W

18W

239W

97%

97%

98%

234W

Overall Efficiency: 97.5%



Sub-module Level MPPTIncreasing Tracking Granularity

Each PV module typically consists of three sub-modulesPV module power output impaired by sub-module mismatch

Junction BoxV+V-

Bypass diodes

Junction BoxV+V-

Bypass diodes

Junction BoxV+V-

Bypass diodes 0 0.5 1 1.5 2 2.5 30

10

20

30

40

50

Current (A)

Pow

er (

W)

Sub−Module1

Sub−Module2

Sub−Module3

Module



Related Work

Generationcontrolcircuit1

Distributedconverters(E to E VP)2

SubMIC(E to B VP)3

This Work

Topology Multi-stagechopper

ResonantSC

Flyback Buck-Boost

Local currentsensing

No No No No

Communication Yes No No Yes

Switch voltagerating

String volt-age

Sub-modulevoltage

String/Sub-modulevoltage

Twice Sub-modulevoltage

MPPT True MPPT Near MPPT Near MPPT True MPPT

Easily Scalable No Yes Yes Yes

Module Inte-grated

No Yes No Yes

1T. Shimizu, et al, “Generation control circuit for photovoltaic modules”.

2J. Stauth, et al, “Resonant switched-capacitor converters for sub-module distributed photovoltaic power

management”.3

C. Olalla, et al, ”Architectures and control of submodule integrated dc-dc converters for photovoltaicapplications”.S. Qin, R. Pilawa-Podgurski (UIUC) Sub-Module DPP for PV Applications 2013.04.01 8 / 23

Proposed Approach

Outline





Proposed Approach Architecture

Sub-module DPP system

DPP Converters implemented as buck-boostEach converter process only power mismatch

A

B

C

Integrated Power Stage

L

Csum

OverlapControl

PWM

Vcin

5VReg

Micro-controllerPWM

R1

R2

Vsum

Vsum

5V

DPP Converter

VDD

VSS

Cmid

Cf1Vmid

Vmid

R3

R4

Cf2

Central

Converter

IstringModule

Sub-

Module

DPP

Converter

DPP

ConverterVstring

A

B

C

1

Sub-

Module

Sub-

Module

2

3



Converter Miniaturization

Low converter voltage andpower ratingLow voltage, high frequencytransistorIntegration into modulejunction box

Central

Converter

IstringModule

Sub-

Module

DPP

Converter

DPP

ConverterVstring

A

B

C

1

Sub-

Module

Sub-

Module

2

3



DPP Converter Implementation

Table: Converter Specifications

Converter Topology Buck-BoostSub-module Voltage Range 3-13.5 VConverter Power Rating 60 WSwitching Frequency 250 kHzConverter Peak Efficiency 97%Converter Weight 4.55 gramsConverter Volume 1.91 cm3


Proposed Approach Control

Control Problem Statement - True MPPT

Pstring = Vstring × Istring

Overall MPP achieved onlywhen MPP at each sub-module(all circuit variables fullydetermined)Central converter performsP&O to Istring in a ‘slow’ loopMaximize Vstring to maximizeoutput powerDPP converters perform P&Oto D1,D2,D3...... to maximizeVstring in a ‘fast’ loop


Proposed Approach Control

System Model

Vstring is a function of dutyratio D1,D2,D3...... givenother parameters fixedThis function only has onemaximum point for a givenstring currentPerturb each duty ratio inturns and observe changein string voltage to find:

arg maxDi∈〈0,1〉

Vstring(Di)3-dimensional plot of string voltage (Vstring )

versus DPP converter duty ratios (D1,D2) fora 3-sub-module 2-converter system


Experiment

Outline





Experiment Experiment Setup

PV Module Emulation1

PV module emulation enables controllable and repeatable PVexperiment in an indoor environment.

HP 6632A

DC Power Supply

To Load

(c) Hardware connection.

RsRp

Id

I

+

PV Module

Ip

-V+

-Vd

Iph=0

Iext

(d) Equivalent circuit schematic.

Figure: Emulation of PV module.

1S. Qin, et al, "Laboratory Emulation of a Photovoltaic Module forControllable Insolation and Realistic Dynamic Performance", PECI 2013


Experiment Experiment Setup

Experiment Setup1

Test system consists of three PV sub-modulesTwo converters communicate to a computer through I2CAn electronic load acts as the central inverter

Iph

Sub-Module 2

Iext2Vsub

+

-

Isub

Dbypass

Iph

Sub-Module 3

Iext3Vsub

+

-

Isub

Dbypass

Iph

Sub-Module 1

Iext1Vsub

+

-

Isub

Dbypass

DPP

DPP

Istring

ElectronicLoad

Schematic drawing of laboratory test setup

0 2 4 6 8 10 12 14Voltage [V]

0

5

10

15

20

25

30

35

Power [W]

Sub-module CharacterizationSC Current (Iext)

0.3A0.6A0.9A1.2A1.5A1.8A2.1A2.4A2.7A3.0A

Power versus voltagecharacteristics of controlled

laboratory PV sub-module setup1S. Qin, et al, "Laboratory Emulation of a Photovoltaic Module for

Controllable Insolation and Realistic Dynamic Performance", PECI 2013S. Qin, R. Pilawa-Podgurski (UIUC) Sub-Module DPP for PV Applications 2013.04.01 17 / 23

Experiment Result and Analysis

Experiment Result

Output current sweep performed by electronic loadSubstantial power output improvementElimination of local maxima caused by bypass diode

0.0 0.5 1.0 1.5 2.0 2.5 3.0Current [A]

0

10

20

30

40

50

60

70

80

Power [W

]

Max Module Power without DPP: 47.3 WMax Module Power with DPP: 60.1 WImprovement with DPP: 27.2 %

Isc1=3A, Isc2=2.1A, Isc3=1.5A

Sub-module 1Sub-module 2Sub-module 3Without DPPWith DPP_MPPT

Moderate mismatch

0.0 0.5 1.0 1.5 2.0 2.5 3.0Current [A]

0

10

20

30

40

50

60

Power [W

]

Max Module Power without DPP: 28.0 WMax Module Power with DPP: 46.2 WImprovement with DPP: 64.9 %

Isc1=3A, Isc2=1.2A, Isc3=0.9A

Sub-module 1Sub-module 2Sub-module 3Without DPPWith DPP_MPPT

Severe mismatch



Experiment Result

Power output improvement in different partial shading conditions

40 50 60 70 80 90 100Insolation Level of the 3rd Sub-module [%]

−10

0

10

20

30

40

50

Impr

ovem

ent w

ith D

PP [%

]Power Output Improvement with DPP

Insolation Level of the 2nd Sub-module40%60%

80%100%



Efficiency Considerations

0.0 0.5 1.0 1.5 2.0Output Current [A]

65

70

75

80

85

90

95

100

Efficiency [%

]

Efficiency of DPP Converter

Fixed Frequency CCMBurst Mode OperationDual Mode Operation


Summary

Summary

Differential power processing can significantly improve poweryield, and reduce size and cost of PV systemTrue MPPT can be achieved with no local current sensingand fast tracking by the architecture and control schemepresentedThe effectiveness of the proposed solution has beenexperimentally verified by a controllable indoor test setup


Summary

QUESTIONS?

Contact:Shibin QinEmail:[email protected]


Summary

Power Rating of DPP Converters


Documents

Sub-Module Differential Power Processing for Photovoltaic ...cemeold.ece.illinois.edu/seminars/Spring13_4_1... · PV module emulation enables controllable and repeatable PV experiment