SAYURI–PV 2016

IEC standardization of IEC standardization of BD Thermal Runaway TestBD Thermal Runaway Test

( IEC62979CD / Doc 82/1025/CD)( IEC62979CD / Doc 82/1025/CD)( IEC62979CD / Doc. 82/1025/CD)( IEC62979CD / Doc. 82/1025/CD)

October 5 2016October 5 2016October 5, 2016October 5, 2016

YasunoriYasunori UchidaUchidaPOWER TECHNOLOGY TESTING LABORATORYPOWER TECHNOLOGY TESTING LABORATORYJAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES (JET)JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES (JET)JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES (JET)JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES (JET)

2

IEC standardization of IEC standardization of BD Thermal Runaway TestBD Thermal Runaway Test

➊ Proposed back-ground

➋ Test conditions of “Thermal Runaway Test”

➌ Current situation of “IEC standardization”

➍ Another effort (Tj measurement)➍ Another effort (Tj measurement)

➎ A k l d t➎ Acknowledgement

3

Quality Assurance Forum in order to establish the high reliability PV module started on July 2011 in San-Francisco US.

TG-1 ; PV QA Guideline for Manufacturing Consistency

TG-2 ; Thermal and Mechanical fatigue including vibration

TG-3 ; Humidity, Temperature and Voltage

TGTG--4 ; Diode, Shading & Reverse bias / Leader Dr. Paul Robust4 ; Diode, Shading & Reverse bias / Leader Dr. Paul RobustDrDr VivekVivek GadeGadeDr. Dr. VivekVivek GadeGade

TG-5 ; UV, Temperature and HumidityTG 5 ; UV, Temperature and Humidity

4

ProgressProgressOct., 2011 Started the activities as Japanese leader of QAF TG4.Oct 2014 NP /New work item Proposal was approved

gg

Oct., 2014 NP /New work item Proposal was approved. Participation : US, DE, CN, CR, JP

Sep., 2015 WD / Working Draft was proposed.Sep., 2015 WD / Working Draft was proposed.Circulated CD for 3 months.

Dec., 2015 Received 44 comments from CN, DE, NL, PT, US, , , , ,Technical 23, Editorial 15, General 6

Feb., 2016 Participated in NREL Reliability PV Module Workshop“Behavior of BD during Thermal Runaway Test”

May ,2016 WG2 Taipei meetingDiscussion about test conditions and pass criteria.

Jun. 15, 2016 sent the CDV draft to WG2 convener.A 26 2016 CD i i b i l i il 18 N 2016Aug. 26, 2016 CD voting is been circulating until 18 Nov. 2016.

Back ground5

Back-groundIn many casesIn many cases,

Bypass Diode (BD)IEC61215 Ed.1 P/N diode

IEC61215 Ed 2 SBD (Schottky Barrier Diode)IEC61215 Ed.2 SBD (Schottky Barrier Diode)(2005/April) ➜ The demerit of SBD is a low

ith t di f ltwithstanding of reverse voltage at the high temperature.

6

• In order to satisfy the requirements of “10.18 Bypass y q ypdiode thermal test” in IEC61215 Ed.2, the bypass diode has been switched to the SBD with low Vf.

• However the breakdown voltage of the reverse bias at• However, the breakdown voltage of the reverse bias at high temperature of the SBD as compared with Si P/N diode is lowdiode is low.

• Therefore, we have to verify the thermal design by the thermal runaway test.

The need for the thermal runaway test7

The need for the thermal runaway test• When the PV module’s surface is in the shade (ex.: shade by the 

protuberance such as utility poles, buildings, and/or leaves) during a the sunny irradiation condition, the current flows to the bypass diode (BD) in the J box through the cell string(BD) in the J‐box through the cell‐string.

• Under these circumstances, the temperature of the BD increases after a certain period of timea certain period of time.

• If the shade suddenly disappears, the reverse bias voltage from the cell’s string is applied to the BD.cell s string is applied to the BD.

• In the situation where thermal runaway might occur after the increasing of the diode leak‐current, when the BD which has the low g ,withstanding reverse voltage at high temperature is used. 

• Therefore, the thermal design of the BD in the J‐box should be verified with the thermal runaway test.

Presented at NREL work‐shop 2012

Typical SBD Typical Si P/N diode

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 8

TaskTask‐‐4     Region US4     Region USDestructive Destructive analysis of the diode of leakage state analysis of the diode of leakage state bb th lth lby by thermalthermal‐‐runaway runaway 

Appearance of Thermal runaway diode X iAppearance of Thermal‐runaway diode → No problem

X‐ray image→ No problem Image to removed the resin

Melting traces of chip→ Typical failure pattern by thermal runaway→ Typical failure pattern by thermal runawayThe property was in a short state.

10

TC82/WG2 Taipei meeting / May, 2016

Five items to be discussed➋-1 Module temperature Approved➋ 1 Module temperature Approved

➋-2 Test conditions of applying of forward current Approved➋ 2 Test conditions of applying of forward current Approved

➋-3 Switching over time to reverse bias voltage Approved➋ g g pp

➋-4 Pass criteria Modified

➋-5 Test target of P/N diode Approved

Test procedure of Thermal runaway test

11

Test procedure of Thermal runaway test

➊ Measurement of initial Ir (reverse current)➊ Measurement of initial Ir (reverse current)➋ Apply a forward current to BD at 75℃ or 90℃.

➌ Apply a reverse bias voltage to BD ithi 10 ft di ti f f d twithin 10ms after disconnection of forward current.

VR

➍ Measurement of Ir after testing 10ms

re o

f Tle

ad ➋

Tem

pera

tur

➌T

Time

12

➋-1 Module temperature

a) 75℃ for Open racka) 75℃ for Open-rack

b) 90℃ for Roof-mount

Based on the outdoor exposure data of NEDO project, 90℃was proposed as the temperature of the roof-mount.After that, 75℃ was added as temperature for the open-mount.

Foundation data / RoofFoundation data / Roof--mountmount

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 13

14

➋-2 Applying of forward current

a) Forward currenta) Forward current

1.25 X Isc (A) of PV module➜ F ll d th t t diti f “B Di d Th l T t”➜ Followed the test conditions of “Bypass Diode Thermal Test”.

b) Applying timeb) Applying time

Apply the forward current of the BD for at least 40 minutes and until the range of Tlead change duringminutes and until the range of Tlead change during 10 minutes becomes within 0.3℃.

15

➋-3 Applying of reverse voltage

Shut off the forward currentShut-off the forward current. Within 10 ms, apply the reverse bias voltage to the BD to be tested.

To verify the effect of the switching time, the energy i di t l ft thof

BD

switching time

Tlead immediately after the reverse bias voltage application was measured by changing of the

Pf Pr

Volta

ge

tem

p. o Tlead

measured by changing of the switching time.⇔

V

Lead

If applyingev

. bia

s

Time ➞

Re

Pf (W) = Forward direction energy just before applying of reverse bias voltage.Pr (W) = Energy immediately after reverse bias voltage was applied.

16

( ) gy y g pp

P W

The effect of Switching over timeThe effect of Switching over time

I.D. # IFPr, W

5 ms 10 ms 30 ms 100 ms 200 ms

9A 0.89 0.89 0.87 0.83 0.78BD with potting

9A 0.89 0.89 0.87 0.83 0.78

11A 2.05 2.05 2.01 1.90 1.75

Note: Pf = 2.61W at 9APf = 3.21W at 11A

With the increase of the switching over time (from 5ms to 200ms), the value of Pr becomes smaller.

This means, Tj will decrease if a longer switching over time is applied.

Th d 10 i id d bl d i bt i bl b PC t ll

17

The proposed 10ms is considered reasonable and is obtainable by PC controller.

Foundation data - - - - - SOT in actual fieldAre switch over times of 10 ms realistic? / Comment ; DE074.5 Test procedure c)Field measurements in real system are desirableField measurements in real system are desirable.

Outdoor measurements were performed.➡ The switching time of 15 ms to 24 ms were obtained (or measured).

Based on the above results, the test condition of 10 ms is proposed.This test condition is achievable by using a high speed response relay.

Typical switch over time (about 22ms)

Position of shading Switch over time, N 4 time (about 22ms)N=4

➊ Left side cluster 20 ～ 24 ms

➋ Center cluster 15 ～ 21 ms

➌ Right side cluster 15 ～ 20 ms

➊ ➋ ➌

➜ M k h di ll b db d ith th d➜ Make shading on a cell by cardboard with thread, shading was cleared by removing the cardboard quickly by hand.

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 18

➋ 4 P F il it i

19

➋-4 Pass or Fail criteriaa) In case that Tlead and leakage current decrease, and if thea) In case that Tlead and leakage current decrease, and if the reverse leakage current at -10V after the test does not increase to more than 5 times of the initial value, the BD is considered safe from the possibility of thermal runaway and pass the test.➜Tlead, Ir が逆ﾊﾞｲｱｽ印加後減少し、Irが初期値に対して、5倍以内であれば合格であれば合格

b) In other case namely if Tlead and leakage current increase or ifb) In other case, namely if Tlead and leakage current increase or if the reverse leakage current at -10V after the test increase to more than 5 times of the initial value, the BD is considered to havethan 5 times of the initial value, the BD is considered to have failed.➜Tlead, Ir が逆ﾊﾞｲｱｽ印加後増加し、Irが初期値に対して、5倍以上であれば不合格

Tlead and leakage current during thermal runaway

20

Tlead and leakage current during thermal runaway

Tlead observation is not considered suitable for the evaluation of thermal runawayTlead observation is not considered suitable for the evaluation of thermal runawayphenomenon.To cope with these difficulties, the observation of leakage current is recommended as aneffective method. The leakage current respond sharply to thermal runaway.g p p y y

熱暴走時の Tleadによる温度上昇は緩やかであるが、漏れ電流のそれは素早く上昇する。

Foundation data - - - - -In case of thermal runaway occurrence, Ir becomes large leakage current. y , g g

The reverse characteristic after thermal runaway shows obvious changes.

Junction-box A type Junction-box B type SBD Diode-A

10I [A] 10I [A] 10I [A]

5 5 5

‐5

0‐10 ‐5 0 5 10

V [V]

‐5

0‐10 ‐5 0 5 10

V [V]

‐5

0‐10 ‐5 0 5 10

V [V]

‐10

5

‐10

5

‐10

5

B80S‐1 試験前測定B80S‐1 熱暴走後

Tyco‐2 試験前測定Tyco‐2 熱暴走後

Olivia 試験前測定Olivia 熱暴走後

Before the testAfter the occurrence of thermal runaway

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 21

22

➋-5 Test target of BDThe test specimen which employs P/N diodes as b di d ld b t d f th th lbypass diode could be exempted from the thermal runaway test required herein, because the capacity of P/N diodes to ithstand the re erse bias is s fficientlP/N diodes to withstand the reverse bias is sufficiently high.

高 逆 がP/N diode は、高温逆耐圧が SBD (Schottky Barrier Diode) に比べ十分高いため、この試験の対象から除く。

Foundation data - - - - -P/N diodes’ capability to withstand the reverse voltage at the high temperature is very p y g g p yhigh (about one thousand volts). P/N diode could be exempted from the required thermal runaway test, because its capability to withstand the reverse bias voltage is high enough.

Typical SBD Typical Si P/N diode

capability to withstand the reverse bias voltage is high enough.

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 23

Presented at NREL work‐shop 2012

Discussion of Thermal runaway test

24

Discussion of Thermal runaway test 19 May, 2016

25

TC82/WG2 members of the Taipei meeting whichTC82/WG2 members of the Taipei meeting, which was held in May 2016.

Current situation of IEC standardization26

Current situation of IEC standardization➊ NP /New work item Proposal finished➋ 3 h fi i h d➋ 3 months vote finished➌ WD / Working Draft finished ➍ CD >>> 3 th t➍ CD >>> 3 months comments

Sep.11 ～ Dec.11, 2015 circulatingsent CDV draft to WG2 convener in 15 Jun 2016sent CDV draft to WG2 convener in 15 Jun. 2016

➎ CDV >>> 12 weeks vote Agreement of P-members > 2/3Negative vote < ¼ of total voteNegative vote < ¼ of total vote

Circulating Aug. 26 – Closing date Nov. 18➏ FDIS preparation by organizer➏ FDIS preparation by organizer➐ FDIS preparation by CO CO (Central office)➑ FDIS 8 weeks vote CO➑

➒ IS / International standard CO

➍➍ Another achievementAnother achievementAs the Tj measurement method for Bypass diode,

10.18.3 Procedure 1 ➞ Tlead / case methodIn the case of Tlead/case method, we should consider "10C margin".The test results showed that about 10The test results showed that about 10℃℃ differencedifference may exist between theThe test results showed that about 10The test results showed that about 10℃℃ difference difference may exist between the calculated Tj by “Tlead method” and the real Tj (Tj measured by Vf-Tj method).This difference could not be overlooked It is consideredThis difference could not be overlooked. It is considered

appropriate to revise the pass criteria of the test to havethe pass criteria of the test to have1010℃℃ margin to the diode manufacturer’s maximum junction temperature margin to the diode manufacturer’s maximum junction temperature

ti i th f th “Tl d th d” i dti i th f th “Tl d th d” i drating in the case of the “Tlead method” is used.rating in the case of the “Tlead method” is used.

10 18 4 Procedure 2 ➞ Vf-Tj method10.18.4 Procedure 2 ➞ Vf Tj methodJET would like to be unified to Vf-Tj method, because it is a primary method a primary method for Tj measurement.

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 27

The relation between Tj and Tlead is expressed as below using the parameter of thermal resistance (Rth)the parameter of thermal resistance (Rth).

Tj = Tlead + ( Vf ×If×Rth )Tj Tlead ( Vf ×If×Rth )The value of Rth is usulally supplied by the diode manufacturer.

Axial type diodeThe problem is that the value supplied is not a real Rth, but an apparent valueof Rth which will vary according to

Axial type diode

of Rth which will vary according to the heat dissipating condition where the diode is installed. (see next slide)

28

B : body

Heat flow and Heat flow and RthRthB : body

✪ presented at NREL PVMR Work-shop 2013 ✪Thermal resistance varies by the difference of heat dissipating condition such as a J-box.

CRc-1 Rc-2A : anode

C : cathode

Cc : heat flow rate on the cathode side

Ca : heat flow rate on the anode side

Th l

AAmbient

Ra-1 Ra-2

Ca : heat flow rate on the anode side

Thermal source (Diode junction) P = Vf X If BRb-1 Rb-2

Cb : heat flow rate on the diode body

Apparent Rth will vary according to the heat flow ratio which varies with the radiation condition of each route !!!!!

Tj = Tlead + Vf×If × Cc × Rth(real) → RthRth(apparent)(apparent)

Tj values indicated below show the measured results and the difference between “Tj1” and “Tj2”.

Kind of BD Tj Tj Tj – TjKind of BD Tj1 Tj2 Tj2 – Tj1

BD-1 / axial 151.7℃ 159.8℃ 8.1℃151.7 159.8 8.1

BD-3 / axial 150.2℃ 162.1℃ 11.9℃

BD-4 / axial 151.8℃ 157.7℃ 5.9℃

BD-5 / TO-220 100.8℃ 108.5℃ 7.7℃

30The test results showed the difference about 10℃.

In “Vf-Tj method” it is necessary to obtain first the "Vf-TjIn Vf-Tj method it is necessary to obtain first the Vf-Tjcharacteristics” which is practically linear as shown below.The formula to show the characteristics will change by an g yindividual diode and the forward current applied.

A l f Vf Tj h t i tiAn example of Vf-Tj characteristics:

Tj = -885 52 X Vf + 407 95Vf-Tj characteristic (at If=9A) of a BD

℃Tj = -885.52 X Vf + 407.95

Tem

p. ℃

31

Vf, V

Conclusion

Aft l i L hb h tiAfter my proposals in Loughborough meeting (Spring 2015),

Procedure 1 (Tj method) was rejected from IECProcedure 1 (Tj method) was rejected from IEC 61215 and/or 61646.

Procedure 2 (Vf Tj method) was remained inProcedure 2 (Vf-Tj method) was remained in IEC 61215 and/or 61646.

Copyright© JAPAN ELECTRICAL SAFETY & ENVIRONMENT TECHNOLOGY LABORATORIES. All Rights Reserved 32

➎ A k l d t

33

➎ Acknowledgement

This work was performed in cooperation with ONAMBAONAMBA Sanken ElectricSanken Electric SOMA OpticsSOMA OpticsONAMBA, ONAMBA, Sanken Electric, Sanken Electric, SOMA Optics, SOMA Optics, Kyocera and SHARPKyocera and SHARP..yy

I would particularly like to thankDr. Paul Robust, Dr. Dr. Paul Robust, Dr. VivekVivek GadeGade andandDr. Kent Whitfield.Dr. Kent Whitfield.Dr. Kent Whitfield.Dr. Kent Whitfield.

34

Thank you for your attention.Thank you for your attention.y yy y