IoT as a key dif­fer­en­tia­tor in the re­new­able en­ergy sec­tor, by Prakash Nayak, Se­nior Fel­low at IET and Di­rec­tor, PEnA Power.

Power Watch India - - CONTENTS - By Prakash Nayak The au­thor is Se­nior Fel­low at IET and Di­rec­tor, PEnA Power.

With the aware­ness about global warm­ing and de­plet­ing of fos­sil fu­els be­ing on the rise, the world is shift­ing to­wards al­ter­na­tive sources of fuel to power its needs. Of them, the most sus­tain­able op­tion is the re­new­able en­ergy. As on June 2016, In­dia’s to­tal power gen­er­a­tion ca­pac­ity touched 303GW which in­cludes 42,85GW of Re­new­able En­ergy. Re­new­able elec­tric­ity sources are tar­geted to grow mas­sively by 2022 to 175GW, in­clud­ing more than the dou­ble of In­dia’s large wind power ca­pac­ity from 27.5GW to 60GW and an al­most 15 fold in­crease in so­lar power from 8.06GW to 100GW and 10 GW from bio power and 5GW from small hy­dro power. The am­bi­tious tar­gets would see In­dia quickly be­com­ing one of the lead­ing green en­ergy pro­duc­ers in the world and sur­pass­ing nu­mer­ous de­vel­oped coun­tries.

The Re­new­able en­ergy field is spread­ing across ge­ogra­phies. It is fast be­com­ing one of the most thriv­ing busi­nesses on the planet. With strong mar­ket, comes stiff com­pe­ti­tion. Hence, ef­fi­ciency, op­er­at­ing costs, prof­itabil­ity, and pro­duc­tiv­ity of course re­li­a­bil­ity are of prime im­por­tance. This is not pos­si­ble by sim­ply fo­cus­ing on me­chan­i­cal or elec­tri­cal engi­neer­ing as­pects. The abil­ity to con­tin­u­ously mon­i­tor and main­tain high per­for­mance real time data over the time will en­hance the de­ploy­ment more ag­gres­sively than be­fore. Also, most of th­ese re­sources be­ing vari­able and dis­trib­uted ge­o­graph­i­cally make it dif­fi­cult to mon­i­tor, main­tain and up-keep the plant. This is where IoT kicks in.

Imag­ine the world where you can not only gen­er­ate and use your own elec­tric­ity, large enough to meet your power re­quire­ment with light­ing, air con­di­tion­ers, hot wa­ter show­ers, etc. Also see the sit­u­a­tion where you can sell ex­cess power to the util­i­ties to earn money. This is what re­new­able en­ergy of­fers. The process, how­ever, isn’t that easy to han­dle. Be it a hy­dro-power plant small enough to run on the stream or creek that flows near your home or a small wind power tur­bine gen­er­a­tor in your farm, in­stalled near your ru­ral res­i­dence or a so­lar plant on roof of your home- it all re­quires nu­mer­ous vari­ables that need be mon­i­tored and main­tained re­sult­ing in op­ti­mised power source. This process will prove to be ex­tremely dif­fi­cult for man­ual labour to per­form ef­fi­ciently. Be­cause, for all th­ese, the source of elec­tric­ity is vari­able. At times it will be sunny and at other times cloudy. It is windy on a par­tic­u­lar day and no wind at all the next day. It is im­pos­si­ble to have ded­i­cated com­puter for ev­ery in­di­vid­ual com­po­nent of the sys­tem when they are dis­trib­uted. Just imag­ine, if we can pre­dict this vari­able na­ture of re­sources in much more sci­en­tific and an­a­lyt­i­cal way, we could man­age and op­ti­mise the gen­er­a­tion along with rest of the sources in the Grid. This is where IoT’s po­ten­tial can be un­leashed.

Be it the wind blade stuck up or gear­box worn-out in the wind tur­bine or the cell’s in­di­vid­ual tem­per­a­tures in a so­lar farm goes high, to­day ev­ery imag­in­able com­po­nent can be mon­i­tored and sensed. Data can be col­lected and also be up­loaded to the cloud for pro­cess­ing. Thanks to in­ter­net and dig­i­tal con­nec­tiv­ity, this process is ef­fi­cient and most cost ef­fec­tive when com­pared to a sin­gle ded­i­cated PC/SCADA han­dling all the vari­ables.

The sen­sors and de­vices used in IoT and its im­mac­u­late op­er­a­tions may be small, but make no mis­take. It is a huge in­dus­try which is de­vel­op­ing and its im­mea­sur­able po­ten­tial will touch ev­ery other sec­tor, most promi­nently in the en­ergy sec­tor, with shift­ing its fo­cus on Re­new­able En­ergy for ex­ploit­ing avail­able nat­u­ral re­sources in abun­dance.

If you are al­ready us­ing cap­tive re­new­able power and are look­ing for IoT’s ap­pli­ca­tions, it is un­der­stood that you know about Pre­dic­tive main­te­nance com­pared to con­ven­tional preven­tive main­te­nance. The pre­dic­tive main­te­nance with man­ual ap­proach, though pow­er­ful as a tool by sens­ing and pre­dic­tive cal­cu­la­tions done with a spe­cific time in­ter­val, is prone to er­rors. Imag­ine a set of sen­sors work­ing like bees round the clock, mea­sur­ing all pa­ram­e­ters ev­ery sec­ond, and per­form­ing pre­dic­tive cal­cu­la­tions and pro­vides a means of pre­dic­tive main­te­nance like no

tech­nol­ogy has ever of­fered. This is where IoT will add value to you and your busi­ness.

From a tech­ni­cal point of view, re­new­able en­ergy sec­tor will be vastly im­proved in the fol­low­ing three ma­jor ways:

Big data an­a­lyt­ics, faster anal­y­sis of sen­sor out­puts and break­ing Up of An­a­lyt­i­cal Data for com­mon man.

When you have so many data to crunch in so less time, so small enough that you can­not pos­si­bly ar­rive at a con­clu­sion in time for an ac­tion, big data is the key. Typ­i­cally in re­new­able en­ergy sec­tor, for an ef­fi­cient op­er­a­tion, a hu­mon­gous amount of data like, his­tor­i­cal me­te­o­ro­log­i­cal in­for­ma­tion, real-time in­for­ma­tion on load pro­file etc., is be­ing trans­ferred and an­a­lyzed to get op­ti­mized out­puts. With the big data anal­y­sis and the cloud stor­age ca­pa­bil­i­ties, com­pa­nies and start-ups in the en­ergy will get to faster re­sults and ul­ti­mately, bet­ter ef­fi­ciency. The com­mon man should be able to un­der­stand the in­for­ma­tion and link them to the real­ity. Like when wind is not blow­ing he can re­late that the gen­er­a­tion is down and he can also pre­dict the gen­er­a­tion from wind tur­bine will be good in th­ese months of the year and also re­late to what hap­pened in pre­vi­ous years. Just imag­ine his mo­bile phone throw­ing up this in­for­ma­tion in­stan­ta­neously. Pre­dic­tive an­a­lyt­ics when used deftly on re­new­able en­ergy power plants can pro­vide ac­cu­rate en­ergy pro­duc­tion fore­casts. The an­a­lyt­ics pre­dicts the ma­chine break­downs or glitches, checks for the cor­re­la­tion of var­i­ous pa­ram­e­ters like ir­ra­di­a­tion, wind speed, tem­per­a­ture, hu­mid­ity, cloud cover, trans­former sta­tus etc. and learns their cause and ef­fect re­la­tion­ship.

De­cen­tral­i­sa­tion of data

One will agree that big data is a must in mass scale de­ploy­ment of re­new­able en­ergy gen­er­a­tion. The down­side of us­ing big data for anal­y­sis is stor­age ca­pa­bil­i­ties, be it the cloud or the hard disk stor­age on site, is very lim­ited as when com­pared to the data that it col­lects and stores. Large data stor­age means large space oc­cu­pa­tion, larger se­cu­rity risk and sub­se­quently the costs that ac­com­pany this also huge. Hence, a de­ci­sion must be made so as to de­cide what data to keep and what to delete. All the es­sen­tial data should be kept for the fu­ture cal­cu­la­tions and us­age trends. The con­ven­tional way to store the said data would be on a cen­tral­ized server, but that is in the past. With the ad­vance­ment in the de­cen­tral­iza­tion of data­base tech­niques, all es­sen­tial data could be stored in in­di­vid­ual data hold­ing cen­tres with en­hanced se­cu­rity and re­li­a­bil­ity.

Self-Learn­ing Ma­chines, Ar­ti­fi­cial In­tel­li­gence Ma­chine Learn­ing

Ac­cord­ing to a study, 20% – 40% of re­new­able en­ergy can­not be used be­cause it is un­sta­ble. Pre­dic­tive an­a­lyt­ics when used deftly on re­new­able en­ergy power plants can pro­vide ac­cu­rate en­ergy pro­duc­tion fore­casts. Hence, this will be used in the pre­dic­tion of the fu­ture load and com­put­ing the real-time big data and stor­age of the said data, the de­ci­sion to be made so as to keep what data and delete the rest can­not be hu­manly done within an ac­cept­able time limit. In due course of time, all the com­put­ing de­ci­sions could be made en­tirely by Ar­ti­fi­cial In­tel­li­gence for most ef­fi­cient op­er­a­tion.

Power sec­tor cap­i­tal­is­ing on IoT

One of the im­por­tant tasks for the In­dian power sec­tor is in­te­gra­tion of re­new­able en­ergy in to the tra­di­tional grid with large part pow­ered from steam tur­bines and hy­dro tur­bines which gen­er­ates power till it is switched off. What is go­ing to be dif­fer­ent from now on­wards when large por­tion of Re­new­able en­ergy 175GW is in­te­grated, which is vari­able, like when wind blows, the ma­chine will gen­er­ate 1000’s of MW and when wind stops the power gen­er­a­tion comes down to few MW in a frac­tion of min/sec, same is the case with so­lar, good sun­shine there is huge gen­er­a­tion sud­den cloud pass­ing through the area it will come down again to few MW. Whereas the load pat­tern may not be vary­ing in this man­ner this will call for much bet­ter pre­dic­tive tool and archived in­for­ma­tion on the be­hav­iours of th­ese sources of en­ergy in ev­ery ge­o­graph­i­cal lo­ca­tions. Also ramp­ing down and ramp up of the base load ma­chines i.e. steam tur­bines and hy­dro tur­bines is also to be done in­stan­ta­neously.This is ex­actly where IoT will play a big role in in­te­grat­ing in most ef­fi­cient and ef­fec­tive way and will be a key dif­fer­en­tia­tor in the re­new­able en­ergy sec­tor and has the po­ten­tial to dra­mat­i­cally trans­form en­ergy sec­tor and op­ti­mis­ing over­all en­ergy mix in the grid.

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.