Update on the Australian National Network PQ Survey

Transcription

1 Update on the Australian National Network PQ Survey Vic Gosbell 1, Sean Elphick 1, Robert Barr 2 1 Integral Energy Power Quality and Reliability Centre, University of Wollongong 2 Electric Power Consulting Contents 1. Introduction 2. Voltage 3. Unbalance 4. Harmonics 5. Sags 6. Customer Severity Index 7. Site Classification Sydney, 26 1

2 1. Introduction Motivation for PQ monitoring 1. Allow utilities to be well informed 2. Find minimum levels that are achievable 3. Determine & prove cost-effective strategies for improvement 4. Motivate a more general understanding of PQ Sydney, 26 2

3 1. Introduction Since 21: Long Term National PQ Survey Long term: full year Utilities use own instruments, including smart tariff meters Disturbances included are - Steady State Voltage - Voltage Unbalance - Harmonics - Voltage Sags Data forwarded to UoW Sydney, 26 3

4 1. Introduction Sites included in UoW PQ Survey Initiatives All sites situated in distribution systems 29 LV 18 MV 28 LV 57 MV 16 LV 93 LV 1 57 MV Sydney, 26 4

5 Outcomes 1. Survey methodology 2. Survey analysis methods 3. Survey reporting 4. Understanding of PQ levels 5. Standards 1. Introduction Sydney, 26 5

6 < % of Sites A. Typical LV histogram 2. Voltage Peak voltage Min limit Typical Voltage Histogram Less than 1 pu More than 1 pu 2.5% 97.5% % 1% Max limit Voltage Spread Voltage (pu) Note limits, V %, V 2.5%, V 97.5%, V 1% Key parameters are - measure of Peak voltage e.g. V 1% - measure of Voltage spread e.g. V 1% -V % Sydney, 26 6

7 >.155 % of Sites B. LV Results - Spread 2. Voltage Histogram of Spread (max - min) Spread (pu) Spread is within 12% band Could be greater at downstream sites! Sydney, 26 7

8 % of Sites 2. Voltage C. LV Results Peak voltage Histogram of Maximum Voltages Voltage (pu) Maximum voltage is higher than expected at many sites Can they be reduced? - downstream measurements needed to clarify Sydney, 26 8

9 > >2.5 % of Sites % of Sites 3. Unbalance LV All Sites Voltage Unbalance MV All Sites Voltage Unbalance % 2% 1% 2% Unbalance (%) Unbalance (%) Apparent anomaly of lower values at LV than MV 1% limit is not achievable at a significant fraction of sites Sydney, 26 9

10 % of Limit 4. Harmonics Trends in average Harmonic Indices Year 1 Year 2 Year 3 Limit 2 LV average MV average LV & MV THD about 4% of limit values LV increases at.1%/annum consistent with overseas reports MV increases at about three times the LV rate Sydney, 26 1

11 PU Sag Voltage 5. Sags CBEMA Curve Protection Curve time seconds CBEMA curve is unsuitable as a measure of distribution utility performance Protection curve gives a better idea of utility capabilities Sydney, 26 11

12 % of Limit 5. Sags Trends in average Sag Indices Year 1 Year 2 Year 3 Limit 2 LV average MV average Decrease in line with severe weather events Lack of strict annual repeatability requires many years of data to establish reasonable utility limits Sydney, 26 12

13 6. Customer Severity Index Definition Aim to find a single measure for the overall impact of all measured PQ disturbances on customers Steps 1. Need a common comparison: normalise all indices relative to maximum acceptable levels 2. Assume V, U, H impact is a thermal effect, varies as square 3. Assume S impact is linear 4. Then CSI =.25 (V 2 + U 2 + H 2 + S 1 ) V: Voltage index U: Unbalance index H: Harmonic THD index S: Sag index Sydney, 26 13

14 CSI LV Result 6. Customer Severity Index Low Voltage Sites CSI unacceptable deserve investigation.4.2. Generally a continuous trend A few really bad sites seem to be atypical & deserve further investigation Sydney, 26 14

15 A. Basis 7. Site Classification Sites can be classified by Network type City Suburban Rural Load type Residential Commercial Industrial Are there significant differences between the PQ levels in these site types? Sydney, 26 15

16 7. Site Classification B. Comparison for LV sites Find average of V, U, H & S for each site classification Determine overall effect using CSI Show CSI & its components by means of stacked bar graph Network type Load type City Sub Rur Ind Com Res V U H S Sydney, 26 16

17 7. Site Classification B. Comparison for LV sites Network type Load type City Sub Rur Ind Com Res V U H S All classifications have acceptable CSI CSI Variation between Network types is 2:1; between Load types is 1.2:1 - hence Network type is more important classification Voltage is largest contribution to CSI, especially at Residential sites Next largest contribution is Unbalance, especially at Rural sites Sydney, 26 17

18 Highlights 1. Systematic approach to conducting, analysing & reporting complex PQ surveys has been developed 2. Voltage Maximum values seem excessive at some sites Need some weak LV sites before final conclusions Future voltage standards could benefit from statistical considerations 3. Unbalance: 1% limit is not achievable at many sites 4. Harmonics could be an issue in a decade or so Sydney, 26 18

19 Highlights 5. Sags Effort needed in developing & promoting sensible equipment immunity standards More years of data required to establish long term performance 6. Customer Severity Index can give overall PQ disturbance measure allowing simple identification of worst sites 7. Site classification At LV, voltage is the most important issue needing addressing The problem is largely independent of Load Type It depends somewhat on Network type, being significantly less at City sites ***** End ***** Sydney, 26 19