Page 80 - ECLECTIC MARCH-2022 INSIDE PAGES_p2.indd
P. 80
consumption) from IT database is to be fetched for Merging the primary dataset
the same docketed consumers to identify the higher
revenue-generating consumers. Next is to merge all the earlier mentioned data to create
f) VIP Consumers: We have a tag for ‘VIP consumer’ a master database.
against certain consumer numbers which are of high
priority in terms of offering service. With VIP tags from Clustering analysis of master database
the same docketed consumers, priority should be considering the consumer number
given in the analysis.
g) Additional Load/AC Application: From the MASD The K-Means clustering of the master database is to be
system, additional load requirement for the same carried out considering consumer number & number of
docketed consumers in near history, if any, is to clusters has been decided 20% of the count of consumers
be plotted to access load growth & more revenue in the list. However, the smart meters & CT services
generation. With sustained low voltage, we can lose are kept out of this process as the consumer number
the same. doesn’t carry the same locational information. Therefore,
clustering the same will not give a fair idea about the
PROCEDURE OF THE ANALYSIS concentrated zone or pocket.

The result is shown below:

Generation of the report of low FigurFeig7:7.SSccrereeennshshoottooffTaTbalbelaeuauouotuptuptudt adtaataofodfidstirsitcrti-cwt-ise   
voltage calls clustering of lowwisveoclltuasgteercianlglsof low voltage calls 3. Mass

The first step is to download a report of low voltage calls VI. SCORING SYSTEM Appr
by consumer number in a specific time period from call- 4. sourc
centre database to create the primary dataset. AS. CSOcoRrinIgNoGf inSdivYidSuaTl cEoMnsumer
From the following Figure, it is evident that, more then
investment required at WSD, ND, HD & SWD with a higher TShceo•crriitenrFigraefqooufrerniancnykdoiinfvgciatdhlleus ca(lrualsnctkeoirnsngasrbeausgmeivdeeonnrbqeuloawnt:ity), 5. Lette
number of consumers affected while ND, SERD, WSD & 6. Adja
SWD has higher per consumer frequency of calls which The•criteAornivaethrfoaegrqeruacanonkntiistnyugmoftphtuienoinctsliuncsotlnaessrutsmsauermed)mg, ievre(nrabnekliongwb: ased
means investing in those areas can address a greater D. Final ran
number of calls. ➤ • FreARqueCpe/eantdicdtyiiotoinofinncaallalllsosta(ydrae(anYrk/(iNnYg/fNobrfmaosaremtdcaltoacsnslaiqfsisueidafineatdsit0ay/s)1,0)/&1), The final sco
All the other data are gathered from the respective source • Step-V * we
as mentioned earlier. 2000 & distri
➤ • AveVraIPgeconcsounmseurm(Yp/tNiofnorminat cllaasstsifsieudmams 0e/r1) (ranking be 2000*10*
Figure 6: Distribution of low voltage calls in the period The ainbbdoaivsveeidddueoatalnirlastnhakerienqagul.laanrrtaitnygoedf uconnitssumcoenr-swuimseeind)c,olumns to will be gener
between 01/04/2019 to 30/09/2021 find
➤ Repetition in last year (Y/N format classified as 0/1), The final rep
consideration
Fc➤oonrseuxAmaCmp/tapiodlend, fiitonirolanaasctol lnsousaumdmm(eYerr/iNfifsrfoe3qr0mu0eanutcncyiltoas,fsscriaefiplelesdtiistai3os,n0a/vi1ne)rla&agset CAPEX inv
y➤ear isVI‘PYecso’nmseuamnse1r ,(AY/CN/afdodrimtioantacllalosasdifi‘eNdoa’sm0e/a1n)s 0 & VIP optimum num
potential high
bcToehn(es3u+am3b0eo0rv+‘eN1+od0’+em0tae)=ailsn30sa40r.,ethaelnl tahrerasncogreedforcothnesucomnesur-mweirsewiilnl
Fi
Bc.oluSmconrisngtoofif nCdlusintedrisvidual ranking.
CESC doesn
For example, for a consumer if frequency of calls is consumer ba
aB3na, osatehvdeeroranscgotehreeciwonnidlilsvubidmeupgaletinsoecnroarteiendoflfaotsrhteeascucohmnscmulumesetrersrisains3a0((0Sculuumsnteiotrsf, , latitude-long
irnedpiveitdituioalnsicnorlaestofyethaer isco‘Ynesus’mmere+an(sN1o, AofC/caodnsduitmioenr ailnlothaed indexation is
sac‘fNeodlundrossrt’tieemhtsirevse/iinctaTygono,nstaapsl0llurnot&mhofeiVteoaIfrfPbowcilclloioiotnlwylnsbuisnimeungme(3trtyeedp+rroem‘cN3ssk0ooe0f’tomecf+doen1craeasn+vullesms0n)0e*u+,r1et0,bh00ye)c)n=co. oSnt3ncohse0,ein4wdts.erecariatonerrgdee meantime, to
investment w
pocket with multiple consumers & prolonged suffered mitigate max
consumers. calculated sc

80  Wit
high
Fig 8. Consumer classification
   75   76   77   78   79   80   81   82   83   84   85