Adaptive fan control

Con en om his wo k may be used unde he e ms o heC ea i eCommonsA ibu ion 3.0 licence. Any u he dis ibu ion o his wo k mus main ain a ibu ion o he au ho (s) and he i le o he wo k, jou nal ci a ion and DOI. Published unde licence by IOP Publishing L d SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 1 Adap i e an con ol Pe Blasinski1,*, Aleš Rubina1, Olga Rubino á1, Š ěpán Jůza1 1B no Uni e si y o Technology, Facul y o Ci il Enginee ing, Ve eří 331/95, 602 00 B no, Czech Republic *E-mail: Pe .Blasinsk[email p o ec ed] Abs ac . The pape p esen s he possibili y o managemen he an pe o mance in dependence on CO₂ concen a ion. A ma hema ical calcula ion is made om he p edic ion o he elec ic ene gy consump ion o a a ian con ol algo i hm. The mos ad an ageous algo i hm is p og ammed in o he ac ual an con ol sys em, and i s powe consump ion is shown in g aph. Finally, he sa ing o he adap i e con ol is compa ed o he classical con ol o e ime. To p og am he con ol algo i hm, he CO₂ mass balance equa ion is used. The equa ion depends on he inpu alues, namely he numbe o CO₂ sou ces, he oom olume and he supply ai . Due o he in ini ely a iable o an con ol, i is possible o adjus he powe se ing depending on he expec ed de elopmen o he loga i hmic ise o CO₂ concen a ion in he oom. This p edic ion is p og ammed wi h he Visual Basic code and p og am code can be used o he A duino con ol uni , which con ols he EC an powe wi h a 0-10 V signal. Al hough signi ican sa ings o e a sho pe iod o ime can no be expec ed, he en ila ion sys em whe e he an uns o a signi ican pa o he yea will be subs an ial. The e o e, he esul will be p esen ed by g aphical diag ams o di e en ways o con olling he an pe yea o ope a ion. 1. In oduc ion Ene gy consump ion du ing en ila ion ope a ion can be signi ican ly educed by designing e icien HVAC sys ems wi h low an inpu . The e iciency o en ila ion sys ems is gene ally ela i ely low so a , bu he sa ings po en ial can be inc eased by a sui able an con ol algo i hm. The ene gy consump ion o ans can be signi ican ly educed by ollowing hese h ee s eps. The i s s ep is o easonably dimension he ai exchange in ensi y by educing as much ai as possible and by using e icien ai dis ibu ion. E icien ai duc s educe unnecessa y o e - en ila ion due o he use o ai - igh duc s, espec o ai low p inciples and ai low con ol. Pe haps he mos impo an hing is o educe he low esis ance and he e o e he an p essu e. This is achie able h ough ae odynamic piping design (including op imum engine oom and ise placemen o educe piping leng h), mo e gene ous dimensioning o piping elemen s, and inc eased uni size, bu wi hou o e all o e sizing o he HVAC sys em. I is necessa y o op imize he e iciency o he HVAC sys em (including an, d i e, mo o and a iable speed d i e, o minimize o e all losses in ensu ing he necessa y ai low and p essu e condi ions). O e sizing should be a oided, as he e iciency o he an may dec ease signi ican ly i he combina ion o ai low and deli e y p essu e is no close o he combina ion achie ing he highes SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 2 e iciency. The e iciency o he mo o and d i e can also d op signi ican ly a low loads. The e o e, o e sizing and a iable loads a e key ac o s a ec ing sys em e iciency. These h ee measu es a e much mo e impo an in clima es whe e he e is a need o hea o cool han some use o na u al d i ing o ces. The a icle ocuses on he i s o hese poin s, especially on he con ol o an ope a ion. 2. Desc ip ion o he oom model and he selec ed an ype In he ma hema ical model, he calcula ion o en ila ion is made o he school class oom. Dimensional cha ac e is ics a e oom wid h 21.0 m, oom leng h 13.3 m, oom heigh 3.8 m. Room olume is 1061 m3, maximum numbe o pe sons is o y. To in oduce a a iable a e o people in he oom, he pe cen age o people was dis ibu ed acco ding o he dis ibu ion shown on Figu e 1. Figu e 1. Pe cen age dis ibu ion o people in he oom Na u al en ila ion o he oom by windows in il a ion was conside ed. Howe e , his pa o he en ila ion componen is minimal. A adial an is conside ed. T adi ionally, he impelle is embedded in a spi al casing ha p o ides an ene gy e icien con e sion o he kine ic ene gy o he lowing ai in o a p essu ized ai . In connec ion wi h he applica ion o new EC mo o s, we can obse e he expansion o adial ans wi h “ ee impelle ”, whe e he spi al box is eplaced di ec ly by he ai -handling uni chambe . This modi ica ion o he an-mo o se signi ican ly educes and cheape , bu a he cos o ae odynamic p ope ies. The ai low in he spi al housing is shown on Figu e 2. [1] Figu e 2. Ai low h ough adial an wi h spi al housing [1] 3. Compu a ional model and desc ip ion o a ian s The compu a ional model o he oom is conside ed as an analy ical model. [2] 𝑉󰇗∙𝑘𝑒𝑑𝜏+𝑀󰇗𝑠𝑑𝜏−𝑉󰇗∙𝑘𝑖𝑑𝜏=𝑂𝑑𝑘𝑖 (1) Whe e: 𝑉󰇗∙𝑘𝑒𝑑𝜏 is he mass o he pollu an in oduced in o he oom by he ai equi ed o en ila ion o e ime dτ 𝑀󰇗𝑠𝑑𝜏 is he mass o he pollu an om he sou ce in he oom o e ime dτ impelle SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 3 𝑉󰇗∙𝑘𝑖𝑑𝜏 is he mass o he pollu an aken om he oom by he ai equi ed o en ila ion o e ime dτ 𝑂𝑑𝑘𝑖 is he weigh gain o he pollu an in he oom ai 𝑘𝑖𝑛𝑡𝑖=𝑘𝑒𝑥𝑡 +(𝑘𝑖𝑛𝑡𝑖−1 −𝑘𝑒𝑥𝑡)∙exp(−𝑉(𝑖−1)∙∆𝜏𝑖 𝑂)+( 𝑀𝑠𝑖 𝑉(𝑖−1))∙(1−exp(−𝑉(𝑖−1)∙∆𝜏𝑖 𝑂)) (2) Whe e: 𝑘𝑒𝑥𝑡 is ou doo CO₂ concen a ion 𝑘𝑖𝑛𝑡𝑖 is maximum CO₂ concen a ion 𝑀󰇗𝑠 is he mass o he pollu an (CO₂) 𝑉(𝑖−1) is o ced en ila ion (es ima ed) 𝑂 is he olume o he oom ∆τ is calcula ion ime in e al Desc ip ion o indi idual an con ol a ian s: Va ian 1 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm, ype o egula ion: on / o . Va ian 2 - minimum an s a in e al is 30 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm, ype o egula ion: on / o Va ian 3 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm and swi ched o only a e eaching a CO₂ concen a ion o 600 ppm, ype o egula ion: on / o Va ian 4 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm and swi ched o only a e eaching a CO₂ concen a ion o 600 ppm, ype o con ol: s epping (2 ans egula ed on / o ) Va ian 5 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm and only swi ched o a e eaching a CO₂ concen a ion o 600 ppm, ype o egula ion: con inuous egula ion (ai low is con olled linea ly acco ding o CO₂ concen a ion) Va ian 6 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm and u ns o a e eaching a CO₂ concen a ion o 600 ppm, ype o egula ion: con inuous egula ion (ai low is con olled quad a ically - con ex acco ding o CO₂ concen a ion) Va ian 7 - minimum an s a in e al is 5 min, he an is swi ched on a a measu ed o CO₂ concen a ion 900 ppm and swi ched o only a e eaching a CO₂ concen a ion o 600 ppm, ype o egula ion: con inuous egula ion (ai low is con olled quad a ically - conca e acco ding o CO₂ concen a ion) Figu e 3. G aphical dependence o an powe con ol, om le : linea , quad a ic - con ex, quad a ic – conca e SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 4 Figu e 3 shows he ai low dependencies on CO2 concen a ion which a e used in he calcula ion o a ian s 5, 6 and 7. These dependencies a ec he s a -up o he an ope a ion and hey signi ican ly in luence he an powe ou pu . 4. Resul s The esul s a e p esen ed by using he g aphs. On igu e 4 he e is a g aph o inc easing CO₂ concen a ion he an ope a ion a ian 1. Figu e 4. G aphical ep esen a ion o he cou se o CO₂ concen a ion educ ion, wi h he an ope a ion a ian 1 When he concen a ion ises abo e 900 ppm, a an is igge ed (ligh blue colo ). The esul s shown ep esen an on/o con ol a ian wi h a minimum ope a ing in e al o 5 min. The g aph shows equen swi ching o en ila ion. G aph on he igu e 5 ep esen s he adap a ion o he an powe o he quad a ic dependence on he CO₂ concen a ion. The en ila o is ope a ing con inuously, bu he powe adap s o he ac ual CO₂ concen a ion. Figu e 5. G aphical ep esen a ion o he cou se o CO₂ concen a ion educ ion, wi h he an ope a ion a ian 7 On he igu e 6 he e a e esul s o he cumula i e CO₂ concen a ions o he an con ol a ian s 1-7. SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 5 Figu e 6. G aphical ep esen a ion o he cou se o CO₂ concen a ion educ ion, wi h he an ope a ion a ian 1-7 The o e all esul s o he an ope a ion pe day, mon h and yea acco ding o he model abo e a e shown in he igu e 7. Figu e 7. Fan ope a ing ime o ope a ing a ian s 1-7 The economic e alua ion is shown on igu e 8. The p ice o 0.17 EUR o kWh o elec ici y was conside ed in he p epa a ion o his cha . The powe inpu o he ans was de e mined om he cha ac e is ics o he an manu ac u e . SEED 2019 IOP Con . Se ies: Ea h and En i onmen al Science 642 (2021) 012008 IOP Publishing doi:10.1088/1755-1315/642/1/012008 6 Figu e 8. P ice pe yea o an ope a ion o each a ian and pe cen age sa ings compa ed o he mos expensi e a ian 5. Conclusion The esul s show ha he g ea es sa ings a e achie ed in a ian 6, which is he con ol o he an acco ding o he quad a ic dependence in he con ex shape. Howe e , i should be added ha his me hod was he only me hod o exceed he ecommended CO₂ concen a ion. This is because he an's s a -up is e y slow in na u e and he an s a s o un a highe CO₂ concen a ions a highe powe . The comp omise be ween pe o mance, sa ings and CO₂ concen a ion esul s om a ian 7, an con ol acco ding o quad a ic dependence in conca e shape. In his case, he en ila o can e ec i ely en ila e he oom and he cos o unning he an is only 13% highe han he mos economical op ion 7. The conclusion is made o one si ua ion (occu ence o people, oom size, e c.). Gi en he na u e o he p oblem o inc easing concen a ion and eac ion o he HVAC sys em, i can be assumed ha e en i he an ope a ing hou s will a y, he a io be ween he di e en a ian s will also change. The o al cos pe yea o an ope a ion conside ed in he calcula ion is no e y high. This si ua ion is due o he ac ha only one an wi h a ela i ely low low a e pe oom is conside ed. Fo buildings such as schools, hospi als, e c., he sa ings would be much mo e signi ican in he sum o all ans. Re e ences [1] DWEYER Tim: Fans o duc ed en ila ion sys ems in Icibse jou nal, online [h ps://www.cibsejou nal.com/cpd/modules/2011-12/] [2] SZÉKYOVÁ, Ma a. Ven ila ion and ai condi ioning. B a isla a: Jaga, 2006, 359 pp. ISBN 80- 807-6037-3. [3] Eu opean S anda d EN 15239. Ven ila ion o buildings – Ene gy pe o mance o buildings – Guidelines o inspec ion o en ila ion sys ems [4] Nilsson, L.J. ‘Ai -handling ene gy e iciency and design p ac ices’, Ene gy & Buildings, Vol. 22 (1995), pp. 1–13 [5] Schild, P. G. ‘Duc less en ila ion – Re u bishing an old lis ed building in o mode n ene gy- e icien o ices’. In Indoo Ai 2008, p oceedings. 2008, pape ID 646. [6] Hydeman, Taylo , S ein, Kolde up & Tong. ‘Ad anced Va iable Ai Volume Sys em Design Guide’, CEC Repo P500-03-082-A-11, Oc obe 2003. [7] Liu, Cla idge & Deng. ‘An ai il e p essu e loss model o an ene gy calcula ion in ai handling uni s’. In . J. o Ene gy Resea ch, Vol. 27 (2003), No. 6, Pp. 589–600 [8] de Almeida, A. T.; P. Ange s; C. U. B unne ; M. Doppelbaue . Mo o s wi h Adjus able Speed D i es: Tes ing P o ocol and E iciency S anda d. 2009