Remote Radio Unit (RRU) DC Feed protection 3 · PDF file1 Remote Radio Unit (RRU) DC Feed 2 protection 3 1. Introduction 4 1.1 RRH Configuration

Document technical content created by M J Maytum

2014 Copyright of M J Maytum

For licence permissions contact [email protected]

i

Remote Radio Unit (RRU) DC Feed 1

protection 2

3




ii

Abstract: Distributed Base Stations (DBS) split the traditional Radio Base Station 1 (RBS) antenna tower base equipment into two locations; a Base Band Unit 2 (BBU) at the tower base and a Remote Radio Unit (RRU) mounted on the top of 3 the tower. Normally the BBU and the RRU would be connected by a fibre optic 4 cable to carry the signals and a DC feed to power the RRU. Towers are likely to 5 be struck with lightning and so some form of protection is necessary to prevent 6 damage to the DC powering feed and connected equipment. Several example 7 protection methods and a worked example are given. Feed cable currents and 8 protection stress levels are calculated for negative and positive lightning flashes. 9 Clause 1 describes the DBS configuration with term definitions in clause 2. 10 Clause 3 shows three possible protection configurations and clause 4 determines 11 the circuit parameters. Clauses 5 through to 7 calculate DC feed cable currents 12 and the Surge Protective Device (SPD) energy for four variants of lightning 13 stroke. Clause 8 outlines two other forms of DC feed protection. Finally clause 9 14 summarises and comments on the results. 15

Keywords: DBS, RRU, RRH, BBU, SPD, DC powering feed, protection, lightning stroke, positive 16 lightning, negative lightning 17

18




iii

Contents 1

1. Introduction ........................................................................................................................................................ 1 2 1.1 RRH Configuration ..................................................................................................................................... 1 3 1.2 History ......................................................................................................................................................... 2 4

2. Definitions and abbreviations............................................................................................................................ 2 5 2.1 Definitions ................................................................................................................................................... 2 6 2.2 Abbreviations .............................................................................................................................................. 3 7

3. Protection configurations ................................................................................................................................... 3 8

4. Circuit component values .................................................................................................................................. 5 9 4.1 Resistances .................................................................................................................................................. 5 10 4.2 Inductances .................................................................................................................................................. 6 11 4.3 Equivalent experimental tower circuit ....................................................................................................... 6 12

5. Tower voltage during a negative lightning flash.............................................................................................. 7 13 5.1 First stroke ................................................................................................................................................... 8 14 5.2 Subsequent strokes .................................................................................................................................... 12 15

6. Tower voltage during a positive lightning stroke .......................................................................................... 14 16 6.1 Positive stroke ........................................................................................................................................... 14 17

7. Tower voltage during an extreme positive lightning stroke .......................................................................... 16 18 7.1 Extreme positive stroke ............................................................................................................................ 17 19

8. Alternative protection arrangements ............................................................................................................... 18 20 8.1 Gas Discharge Tube (GDT) reference bonding ...................................................................................... 18 21 8.2 Shielded DC feed ...................................................................................................................................... 19 22

9. Lightning stroke results and comments .......................................................................................................... 19 23

Annex A (informative) Bibliography................................................................................................................. 23 24

25

26




1

Remote Radio Unit (RRU) DC Feed 1

protection 2

1. Introduction 3

1.1 RRH Configuration 4

Traditionally Radio Base Stations (RBS) had the bulk of the equipment at the base of the antenna tower or 5 mast. Distributed Base Stations (DBS) split the equipment into two sections; a Base Band Unit (BBU) at 6 the tower or mast base and a Remote Radio Unit (RRU) mounted at the top of the tower or mast. Normally 7 the BBU and the RRU would be connected by a fibre optic cable to carry the signals and a DC feed to 8 power the RRU, see Figure 1. The RRU is also called a Remote Radio Head (RRH). To harmonize with 9 [B3] the acronym RRU will be used in this document. 10

Where a tower exists, such as power distribution pylons, the RRU may be mounted at an intermediate level 11 rather than at the tower top. 12




2

Remote RadioUnit RRU

Antennas

Powering: BlueSignal: Yellow

Lightning Rod

Base BandUnit RBU

Tower orMast

13

Figure 1 — Distributed Base Station 14

1.2 History 15

RRU DC feed protection has been under discussion in the ITU-T since the submission of [B1] in 2011. The 16 2012 presentation [B2] gave details the unexpected DC feed protection surge waveshape and the reasons 17 for it. In 2014 the ITU-T published a Recommendation [B3] giving a more comprehensive explanation of 18 the DC feed surge protection. 19

2. Definitions and abbreviations 20

2.1 Definitions 21

For the purposes of this document, the following terms and definitions apply. 22

Radio Base Station (RBS): Installation intended to provide access to the telecommunication system by 23 means of radio waves. 24 [B3] 25

Distributed Base Station (DBS): One kind of Radio Base Station, where the Remote Radio Unit (RRU) 26 and Base Band Unit (BBU) can be installed separated. 27 [B3] 28




3

Remote Radio Unit (RRU): 29 (Syn. Remote Radio Head (RRH)) 30 The radio frequency module of Radio Base Station which can be installed separately. Optical fibre is 31 commonly used to connected radio frequency module and base band unit of Radio Base Station. 32 [B3] 33

Base Band Unit (BBU): The base band module of Radio Base Station which can be installed separately. 34 Optical fibre is commonly used to connected base band module and radio frequency module. 35 [B3] 36

surge reference equalizer: A surge protective device used for connecting equipment to external systems 37 whereby all conductors connected to the protected load are routed, physically and electrically, through a 38 single enclosure with a shared reference point between the input and output ports of each system. 39 [B4] 40

2.2 Abbreviations 41

BBU Base Band Unit 42

DBS Distributed Base Station 43

RBS Radio Base Station 44

RRU Remote Radio Unit 45

RRH Remote Radio Head 46

RTN Return 47

SPC Surge Protective Component 48

SPD Surge Protective Device 49

3. Protection configurations 50

Figure 2 shows the common arrangement where the floating DC feed has protection applied at the RRU 51 and the BBU. During lightning stroke events the protection, the red block in Figure 2, becomes a surge 52 reference equalizer, lightning event bonding the DC feed to the local tower structure via the black 53 connection line from the protection block to the tower. 54




4

Remote RadioUnit RRUProtection

Powering: BlueSignal: Yellow

Lightning Rod

Base BandUnit RBU

Tower orMast

Protection

55

Figure 2 — Protection (red block) applied to DC powering feed at tower base and top 56

Figure 3 through to Figure 5 shows some protection configurations for a -48 V DC feed, assuming the 57 lightning stroke current flows from tower top to bottom. The lightning stroke current, IS, divides between 58 the tower, IT, and DC feed, IR and IF. Because of the values of tower and DC feed impedances, most of the 59 stroke current will flow in the tower. In the Figures, the tower impedances are RT and LT and the DC feed 60 impedances are RR and LR for the return and RF and LF for the -48 V. The mutual inductance between the 61 tower and DC feed is represented by MTRF. 62

-48 V

RTN

SPD4SPD3

SPD2SPD1

RBURRU

TowerStroke

IS IT

RR

RT

LR

LT

MTRF

IR

RF LFIF

MTRF

MTRF

63

Figure 3 — 2 SPD Protection of DC powering feed 64

In Figure 3, SPD1 and SPD2 at the tower top and SPD3 and SPD4 at the base reference equalize the DC 65 feed voltage to the local tower voltage during a lightning stroke. 66




5

SPD1

SPD2

SPD3

-48 V

RTN

SPD4

SPD5

SPD6

RBURRU

TowerStroke

IS IT

RR

RT

LR

LT

MTRF

IR

RF LFIF

MTRF

MTRF

67

Figure 4 — 3 SPD Protection of DC powering feed 68

Figure 4 is similar to Figure 3, but has the addition of an extra SPD, SPD 2 and SPD5, across the DC feed 69 at the tower top and base to limit the differential feed surge voltage. 70

SPD1

SPD2

-48 V

RTN

SPD3

SPD4 RBURRU

TowerStroke

IS IT

RR

RT

LR

LT

MTRF

IR

RF LFIF

MTRF

MTRF

71

Figure 5 — Single SPD for reference equalization 72

Figure 5 uses a single SPD for reference equalization, SPD1 and SPD3, and another SPD, SPD2 and SPD4, 73 across the DC feed at the tower top and base to limit the differential feed surge voltage. 74

4. Circuit component values 75

Reference [B1] measured a three leg, 5 cm square tube, folded 24 m tower in a laboratory environment 76 fitted with a 6 mm2 DC feed cable. 77

4.1 Resistances 78

The three mild steel tower legs had 2 mm walls, making the total cross-sectional area, A, 3 x (502-462) = 79 1152 mm2. The calculated tower resistance will be: 80




6

mA

lRT 1.3

101152

24105.1

6

7 81

Where: 82

l = 24 m 83 A = 1152x10-6 m2 84 = 1.5x10-7 m (mild steel) 85

The actual measured tower resistance was 2.2 m equivalent to 6.6 m per leg. 86

Similarly the 6 mm2 DC feed cable resistance calculates to 68 m using = 0.17x10-7 m (copper). 87

4.2 Inductances 88

To enable short connection leads to the surge generator, the tower was bent back on itself to form a “U” or 89 hairpin shaped structure. Had the tower been straight, the equation for a straight conductor inductance, 90 L=0.2l(ln(2l/r)-0.75) µH, based on [B5] could have been used. This equation results in an inductance value 91 of 32 µH per tower leg or 11 µH with three in parallel. A tower inductance relationship is given in [B6], 92 which gives a rule of thumb of 0.84 µH for every metre of tower height, giving an inductance value of 93 20 µH for a 24 m tower. 94

The inductance of a hairpin shaped tower can be treated as rectangle with sides of length L and width d 95 formed by a conductor of radius r. The inductance equation [B9] for a rectangle is: 96

Hr

ll

r

dd

d

dl

d

ll

l

dl

l

dddldlL

2ln

2lnlnln224.0

22222297

This equation results in an inductance value of 17 µH per tower leg or 5.8 µH with three in parallel. 98

Using the equation for the 6 mm2 DC feed cable gives a value of 32 µH. Bending the tower round in a 99

hairpin to apply the surge generator reduces the tower and cable inductance. 100

The mutual inductance between the tower and cable is neglected as the recorded current waveforms did not 101 show any major mutual inductive effects. 102

4.3 Equivalent experimental tower circuit 103

Figure 6 shows the simplified circuit for [B1] using the values from previous clauses showing currents and 104 component voltages. 105




7

SPD1+

SPD2

RBU

RRUTop

Stroke

LC

IT

32 µH

Cable

LT

5.8 µH

IC

RT RC

68 m2.2 m

VLC

Base

VLT

VRCVRT

VSPD

Tower

IS

Ground 106

Figure 6 — Folded 24 m tower simplified circuit 107

The tower path is lower in impedance than the cable path so most of the current flows in the tower path. As 108 most of the current is in the tower path the tower acts like a voltage generator to drive current through the 109 cable path. The voltage across the parallel arms must have the same value, giving: 110

RCSPDLCRTLT VVVVV 111

Rearranging in terms of VLC 112

RCSPDRTLTLC VVVVV 113

Integrating this equation with time will result in the voltseconds applied to the cable inductance, LC, which, 114 when divided by the cable inductance will give the cable current, IC. This approach is used to determine the 115 cable current in the following clauses. 116

5. Tower voltage during a negative lightning flash 117

The median stroke values from [B7] will be used to emulate a negative lightning flash. The parameter 118 values are: 119

First stroke: 30 kA, 5.5/75 and 5.2 C 120




8

Subsequent strokes: 12 kA, 1.1/32 and 1.4 C 121

Inter-stroke interval: 60 ms 122

Subsequent stroke number: 3 to 5 123

5.1 First stroke 124

Figure 7 shows the simulated first stroke tower current. The slow start to current rise is to model more 125 accurately the naturally occurring stroke initial current rise. A double exponential lightning equation gives 126 a rapid initial rate of rise and overestimates the inductive voltage levels. The alternative equation given in 127 [B8] was used to emulate the current waveform slow initial rise. 128

Time — µs

0 20 40 60 80 100 120

To

wer

Curr

en

t —

A

0

5000

10000

15000

20000

25000

30000

129

Figure 7 — First stroke tower waveform 130




9

Figure 8 shows the tower inductive (VLT) and resistive (VRT) voltage components as a result of the first 131 stroke current. 132

time µ vs i

Time — µs

0 20 40 60 80 100 120

Com

ponent

Volta

ge —

V

-5000

0

5000

10000

15000

20000

25000

30000

35000

40000

Inductive Voltage

Resistive Voltage

133

Figure 8 — First stroke tower voltages 134

The stroke current rate of rise develops a substantial tower inductive voltage (40 kV red line) and this 135 determines the voltage applied to the cable and the consequent current in the cable. The tower resistive 136 voltage developed in this case is negligible (green line). Until the combined voltage thresholds (VSPD) of the 137 two SPDs, SPD1 and SPD2, is exceeded, substantive current cannot flow in the cable. If, for example the 138 combine threshold voltage was 200 V only the portion of the tower voltage exceeding 200 V will build up a 139 current in the cable. 140

Figure 9 shows the portion of the inductive voltage that exceeds 200 V in red and that below 200 V in 141 black. The fill represents the volt-seconds (172 mVs) applied to the cable inductance, LC. 142

Time — µs

0 10 20 30 40 50

Ind

uctiv

e V

oltag

e —

V

-5000

0

5000

10000

15000

20000

25000

30000

35000

40000

< 200 V

> 200 V

143

Figure 9 — First stroke waveform above 200 V 144

Figure 10 integrates the red portion (>200 V) volt-seconds of Figure 9 and shows the cumulative millivolt-145 seconds applied to the cable inductance LC. As the volt-seconds divided by the cable inductance is the value 146 of current, the cumulative waveshape is also the cable current rise waveshape. The total volt-seconds 147




10

applied is 172 mVs and that would give a peak inductive current of 172m/32µ = 5380 A. The inductive 148 cable energy stored is 0.5*32µ*53802 = 460 J. 149

Time — µs

0 5 10 15 20 25 30

Cu

mula

tive

mV

s

0

20

40

60

80

100

120

140

160

180

200

150

Figure 10 — Cumulative millivolt-seconds applied to the cable inductance LC 151

Figure 11 shows the simplified circuit elements just after the cable current peak. 152

SPD1+

SPD2

RBU

RRUTop

Stroke

LC

IT

32 µH

Cable

LT

5.8 µH

IC

RT RC

68 m2.2 m

VLC

Base

VLT

VRCVRT

VSPD

Tower

IS

Ground

5380 A30000 A

153

Figure 11 — Simplified circuit elements for cable current decay 154

After the cable current has peaked, the cable stored inductive energy discharges. The voltage polarity 155 across the cable inductance (VLC) and tower (VLT) reverses after the current peak as the stroke di/dt 156




11

becomes negative. The voltage across the cable inductance will essentially be the combined SPD clamping 157 voltages of SPD1 and SPD2 (200 V) and the tower voltage. The previous voltage equation becomes 158

RCSPDRTLTLC VVVVV 159

The modified equation shows that the combined SPD voltage now adds to the tower voltage instead of 160 subtracting from it, so reducing the time for the cable inductive energy to discharge. 161

Time — µs

25 50 75 100 125 150 175 200

Tow

er

+ S

PD

vo

ltag

e a

fte

r cu

rre

nt p

ea

k —

V

-2500

-2000

-1500

-1000

-500

0

162

Figure 12 — Tower voltage after current peak 163

The tower voltage plus the voltages of the SPD1 and SPD2 is applied to the cable inductance. Figure 12 164 shows the combined tower and SPD voltage after the stroke current peak. By integrating the total voltage 165 with time, the time at which the volt-second product reaches 172 mVs can be found. The 172 mVs value is 166 reached at 168 µs. At 168 µs the induced current in the cable will be zero. 167

Figure 13 shows the tower and cable currents. The cable current drops to zero at 168 µs or 151 µs after the 168 current peak. 169

The SPD energy deposited can be found from integrating the cable current and multiplying by the SPD 170 voltage (200 V). The total SPD energy deposited was found to be 68 J comprised of 5 J in the current rise 171 and 63 J in the current decay. The total energy deposited in each SPD is 34 J. This is less than 10 % of the 172 energy (460 J) that was stored in the cable inductance. 173




12

Time — µs

0 50 100 150 200 250

First

str

oke t

ow

er

curr

ent

— A

0

5000

10000

15000

20000

25000

30000

Cable

Cu

rrent —

A

0

1000

2000

3000

4000

5000

6000

First stroke tower current

Cable current

174

Figure 13 — First stroke tower and cable currents 175

In summary, the significant first stroke parameters were: 176

Peak tower current 30 kA (stroke current 35 A) 177

Peak cable current 5.38 kA 178

Cable volt-second integral during current rise 172 mVs 179

Cable stored energy 460 J 180

Cable current time from peak to zero 161 µs 181

SPD energy 68 J (about 34 J each SPD) 182

5.2 Subsequent strokes 183

Subsequent strokes are typically lower in amplitude, faster rate of current rise and shorter decay times, but 184 they occur in multiples of 3 to 5. Figure 14 shows the simulated subsequent stroke tower current. 185




13

Time — µs

0 10 20 30 40 50 60 70 80 90 100

Subsequ

ent str

oke tow

er

curr

ent

— A

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000

13000

186

Figure 14 — Subsequent stroke tower current 187

Figure 15 shows the tower inductive and resistive voltage components as a result of the subsequent stroke 188 current. 189

Time — µs

0 10 20 30 40 50

To

we

r co

mp

on

en

t vo

ltag

es

— V

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

Inductive voltageResistive voltage

190

Figure 15 — Subsequent stroke tower voltages 191

The fast rate of current rise causes a peak tower voltage of over 80 kV. The treatment of these values 192 followed the same procedure as the first stroke clause. Significant parameters were: 193

Peak tower current 12 kA (stroke current 14 kA) 194







14



Figure 16 shows the tower and cable currents. If there were 5 subsequent strokes the total energy deposited 200 in a single SPD (SPD 1 or SPD2) would be 5*7 + 34 = 69 J. 201

Time — µs

0 10 20 30 40 50 60 70 80

Sub

sequen

t str

oke

tow

er

curr

en

t —

A

0

2000

4000

6000

8000

10000

12000

Cab

le c

urr

ent

— A

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

2400

Subsequent stroke tower current

Cable current

202

Figure 16 — Tower and cable currents 203

6. Tower voltage during a positive lightning stroke 204

The median stroke values from [B7] will be used to emulate a positive lightning stroke. The parameter 205 values are: 206

35 kA, 22/230 and 16 C 207

6.1 Positive stroke 208

Figure 17 shows the simulated positive stroke tower current. The slow start to current rise is to model more 209 accurately the naturally occurring stroke initial current rise. 210




15

Time — µs

0 50 100 150 200 250 300 350 400 450 500

Positiv

e s

troke t

ow

er

curr

ent —

A

0

5000

10000

15000

20000

25000

30000

35000

40000

211

Figure 17 — Positive stroke tower current 212

Figure 15 shows the tower inductive and resistive voltage components as a result of the subsequent stroke 213 current. 214

Time — µs

0 50 100 150 200 250 300 350 400 450 500

To

we

r co

mp

on

en

t vo

ltag

es

— V

-1000

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

11000

12000


215

Figure 18 — Positive stroke tower voltages 216

The rate of current rise causes a peak tower voltage of 12 kV. The treatment of these values follows the 217 same procedure as the first stroke clause. Significant parameters were: 218








16



Figure 16 shows the tower and cable currents. This figure shows that, with the 35 kA stroke current, when 225 the stroke current time to half value is 250 µs or longer the cable current decay waveshape tends to a linear 226 ramp. This was noticed on the oscilloscope shots of [B1] where the surge generator used had a time to half 227 value of 400 µs to 440 µs. 228

Time — µs

0 50 100 150 200 250 300 350 400 450 500

Po

siti

ve s

troke t

ow

er

curr

ent —

A

0

5000

10000

15000

20000

25000

30000

35000

40000

Cab

le c

urr

ent

— A

0

1000

2000

3000

4000

5000

6000

7000

8000

Positive stroke tower current

Cable current

229

Figure 19 — Positive stroke tower and cable currents 230

7. Tower voltage during an extreme positive lightning stroke 231

The recommended extreme (1 % population) stroke values from [B11] were used to emulate an extreme 232 positive lightning stroke. The positive stroke parameter values given are: 233

350 kA peak current, 11 µs time to current peak, 40 µs time to half value and 500 kA/µs di/dt 234 maximum 235

For reference the extreme negative stroke parameter values given are: 236

160 kA peak current, 6 µs time to current peak, 80 µs time to half value and 500 kA/µs di/dt 237 maximum 238

Both waveforms have a 500 kA/µs maximum rate of rise and will generate the same value of peak 239 inductive voltage. 240




17

7.1 Extreme positive stroke 241

Figure 20 shows the simulated extreme positive stroke tower current. This simulation only achieves a 242 maximum di/dt of 100 kA/µs and not the recommended 500 kA/µs. 243

Time — µs

0 20 40 60 80 100 120 140

Extr

em

e p

ositi

ve s

troke tow

er

curr

ent

— A

0

50000

100000

150000

200000

250000

300000

350000

400000

244

Figure 20 — Extreme positive stroke tower current 245

Figure 21 shows the tower inductive and resistive voltage components as a result of the extreme positive 246 stroke current. 247

Time — µs

0 20 40 60 80 100 120 140

To

wer

com

ponent

volta

ges

— V

0

100000

200000

300000

400000

500000

600000

700000


248

Figure 21 — Extreme positive stroke tower voltages 249

The rate of current rise (105 kA/µs) causes a peak tower voltage of 620 kV. The treatment of these values 250 follows the same procedure as the first stroke clause. Significant parameters were: 251






18

Cable volt-second integral during current rise 1.79 Vs 254

Cable stored energy 50 kJ 255



Figure 22 shows the tower and cable currents. The tower inductive voltage dominates making the cable 258 current waveshape similar to the stroke current waveshape. 259

Time — µs

0 20 40 60 80 100 120 140

Tow

er

curr

ent —

A

0

50000

100000

150000

200000

250000

300000

350000

400000

Ca

ble

curr

ent —

A0

10000

20000

30000

40000

50000

60000

70000

80000

Tower current Cable current

260

Figure 22 — Extreme positive stroke tower and cable currents 261

8. Alternative protection arrangements 262

8.1 Gas Discharge Tube (GDT) reference bonding 263

Clause B.5 “Series connected GDTs for DC power applications” in [B10] shows how a series combination 264 of GDTs can develop a combined arc voltage higher than the protected DC supply voltage. This series 265 arrangement avoids the possibility of the GDTs continuing to conduct after a lightning surge. 266

The previous Figure 5 circuit can be used with single GDTs for SPD1 and SPD3. 267




19

SPD1

SPD2

-48 V

RTN

SPD3

SPD4 RBURRU

TowerStroke

IS IT

RR

RT

LR

LT

MTRF

IR

RF LFIF

MTRF

MTRF

268

Figure 23 — Single GDT for reference equalization 269

When the tower stroke voltage causes the sparkover of the GDTs, the return, RTN, cable is bonded to the 270 tower top and base. As a result a portion of the stroke current will flow in the cable RTN for the stroke 271 duration. SPDs SPD2 and SPD4 should be clamping type voltage limiters to avoid shorting the DC supply. 272 The feed (-48 V) cable current will be similar to that of the clause 4.3 circuit. 273

8.2 Shielded DC feed 274

The Appendix I, Isolated protection solution: example of [B3] shows how a cable shield and a power feed 275 isolation barrier in the RRU could be used to remove the need of SPD elements. The withstand voltage of 276 the isolation barrier needs careful consideration to comprehend the maximum stroke voltages. In this 277 document peak values of 80 kV have been calculated just for nominal conditions on the 24 m tower. 278

9. Lightning stroke results and comments 279

Table 1 —Lightning stroke values and results 280

Stroke (clause #)

Peak tower

current kA

Cable peak

current kA

Stroke waveshape

Cable current time to 0

µs

Vs integral

peak mVs

Tower energy

J

Cable energy

J

Total SPD

energy J

Negative first (5.1)

30 (100 %)

5.4 (18 %)

5.5/75 160 170 2610 (100 %)

460 (18 %)

68 (2.6 %)

Negative subsequent (5.2)

12 (100 %)

2.2 (18 %)

1.1/32 70 69 418 (100 %)

75 (18 %)

13 (3.1 %)

Positive (6.1)

35 (100 %)

6.1 (17 %)

22/230 340 200 3550 (100 %)

600 (17 %)

210 (5.9 %)

Extreme positive (7.1)

350 (100 %)

56 (16 %)

4.5/40 110 1800 355000 (100 %)

50000 (14 %)

460 (0.13 %)

Percentage values shown are relative the tower current peak or tower energy




20

Table 1 results were obtained by the following procedure: 281

1) Estimate the inductive and resistive components of the tower and DC feed cable. 282

2) Define the lightning stroke waveform applied to the model. 283

3) Assume most of the stroke current flows in the tower 284

4) Calculate the tower voltage from the stroke current and the tower inductance and resistance. 285

(in this example the resistive voltage was negligible) 286

5) Subtract the nominal SPD voltage from the result of 4). 287

6) Integrate the result of 5) with time to give a cumulative Vs waveform. 288

7) Divide the result of 6) by the cable inductance to give the SPD current versus time waveform. 289

8) Multiply the result of 7) by the nominal SPD voltage to give the SPD power versus time waveform. 290

9) Integrate the result of 8) for the cable current duration time to give the SPD cumulative energy. 291

Only a nominal value of SPD clamping voltage was used. A further refinement of this procedure would be 292 to incorporate a model for the SPD clamping voltage versus current. 293

The lightning parameters used for analysis of the first three table data rows were median stroke values from 294 [B7]. The forth extreme positive stroke row used recommendations from [B11], which represents the more 295 stressful values occurring in the field. For a given amplitude of stroke current, the energy deposited into the 296 SPD increases with increasing time to half value of the stroke. Longer times, over about 250 µs, to half 297 value cause the SPD current decay waveshape to approach a linear ramp. 298

The transition from a truncated stroke waveshape to a linear ramp warrants further investigation. The 299 following (simplistic) approach provides some indication of the transition region. The actual tower current, 300 iT, can be approximated to: 301

TC

CST LLL

ii 302

Where: 303

iS = Peak stroke current 304 LT = Tower inductance 305 LC = Cable inductance 306

In Figure 24 the tower current is approximated to a triangular waveform. The waveform rises in time TF to 307 a peak of current of iTMAX and then linearly decays to 50 % of iTMAX, in time TDT, measured from the current 308 peak. The cable current reaches zero in time 2TDT. 309




21

Voltage

To

we

r C

urr

ent

Cable

Cu

rre

nt

0

0

0

iCMAX

iTMAX

vTMAX

-vSPD

-vSPD

0.5iTMAX

2TDTTF0

TF0

iT actual

iT linear approximationTDT

2TDC

310

Figure 24 — Simple analysis of cable current decay time to zero 311

In the time to peak current, the tower current rate of rise generates an inductive voltage, VTMAX, of 312 LT*iTMAX/TF. The voltage applied to the cable inductance, LC, will be reduced by the SPD clamping voltage 313 VSPD. The peak cable current, iCMAX, will be: 314

CSPDFTMAXTFCMAX LVTiLTi // 315

Simplifying to 316

CSPDFCTMAXTCMAX LVTLiLi // 317

Generally the TFVSPD/LC factor will be small making 318




22

CTMAXTCMAX LiLi / 319

During the tower current decay time the cable inductive voltage, vCD, will be 320

SPDDTTMAXTCD VTiLv 2/ 321

or 322

DTSPDDTTMAXTCD TVTiLv 2/2 323

The time, 2TDC, for the cable current to decay to zero is given by 324

CDCMAXCDC viLT /2 325

Simplifying 326

SPDDTTMAXTTMAXTDTDC VTiLiLTT 2/22 327

TMAXTSPDDTDTDC iLVTTT /21/22 328

The equation above indicates that the cable current decay linearizes when the 2TDTVSPD/(LTiTMAX) factor 329 becomes significant compared to unity. Inserting the example values gives 400TDT/(5.8iCMAX) = 330 69TDT/iTMAX, where TDT is in microseconds. In clause 6.1 with iTMAX = 35 kA and TDT =200 µs a linear 331 ramp is noticeable. The factor for this condition is 69*200/35000 = 0.39, making 2TDC = 2*200/1.39 = 290 332 µs. The actual calculated value using exponential decays was 340 µs. 333

The above shows the transition from a cable current truncated exponential decay waveshape to a linear 334 decay ramp is determined by the 2TDTVSPD/(LTiTMAX) factor becoming significant compared to unity. For a 335 given setup VSPD and LT will be fixed and the variables are TDT and iTMAX. In the example given, a 336 linearized ramp occurred with a factor of TDT/iTMAX = 200 µs/35 kA = 5.7 nAs. 337

Further the approximate SPD energy, ESPD can be calculated with 338

CTDCTMAXSPDSPD LLTiVE / 339

Using previous values ESPD = 200*35000*100 µs*5.8/32 = 20*35*5.8/32 = 126 J. The actual calculated 340 value using exponential decays was 210 J. Thus the simplistic linear approach can be used to establish 341 orders of magnitude but not to any level of accuracy. 342

The analyzed tower was 24 m and folded. Higher towers will result in proportionally larger values of stress 343 as the inductances will be proportional to the tower height. 344




23

Annex A 345 346

(informative) 347 348

Bibliography 349

These bibliographical references provide background information, but do not need to be consulted to 350 understand this document. 351

[B1] Mr Politis Zafiris, Lightning current through MOV based SPDs in RRH applications, Raycap SA 352

[B2] Mick Maytum, Towering Powering protection problem on Remote Radio Head (RRH) cellular 353 systems, PEG 2012 Conference 354

[B3] ITU-T Recommendation K.97 (02/2014): Lightning protection of distributed base stations 355

[B4] IEEE Std 1100-1999 - IEEE Recommended Practice for Powering and Grounding Electronic 356 Equipment 357

[B5] IEEE Std 518-1982 - IEEE Guide for the Installation of Electrical Equipment to Minimize Electrical 358 Noise Inputs to Controllers from External Sources 359

[B6] E. A. Williams, G. A. Jones, D. H. Layer, T. G. Osenkowsky, National Association of Broadcasters 360 Engineering Handbook, Focal Press 361

[B7] CIGRÉ (Council on Large Electric Systems) Technical Bulletin (TB) 549 (2013), Lightning 362 Parameters for Engineering Applications 363

[B8] F. Heidler, Z. Flisowski, W. Zischank, C. Mazzetti, Parameters of lightning current given in IEC 364 62305 – Background, Experience and Outlook, 29th International Conference on Lightning Protection, 365 23rd – 26th June 2008 – Uppsala, Sweden 366

[B9] Rectangular loop inductance, 367 http://www.cvel.clemson.edu/emc/calculators/Inductance_Calculator/rectgl.html, Clemson University, 368 retrieved 2014-10-28 369

[B10] IEEE PC62.42.1 - Draft Guide for the Application of Surge-Protective Components in Surge 370 Protective Devices and Equipment Ports - Gas Discharge Tubes (GDTs) 371

[B11] W. R. Gamerota, J. O. Elisme, M. A. Uman and V. A. Rakov, Current Waveforms for Lightning 372 Simulation, IEEE Transactions on Electromagnetic Compatibility, Vol. 54, No. 4, August 2012, pp 880-888 373

Documents

Remote Radio Unit (RRU) DC Feed protection 3 · PDF file1 Remote Radio Unit (RRU) DC Feed 2 protection 3 1. Introduction 4 1.1 RRH Configuration