THE EFFECTIVENESS OF FIRE DETECTION SYSTEMS IN DIFFERENT DIMENSIONS Agoson RESTAS ÚČINNOSŤ SYSTÉMOV POŽIARNEJ DETEKCIE V RÔZNYCH ROZMEROCH Absrac Inroducion: There are many ways of fire deecion; in many cases aking an overview of he effeciveness is no useless. Mehods: Looking a he moniored fores as a fla area gives a dimension analysis bu as an ariculaed area gives a 3 dimension analysis. If research couns wih he ime scale, meaning he way weaher condiion changes, i gives a 4 dimension analysis. Resuls and discussion: Research says in he dimension model he aerial fligh parol canno be effecive, a ower based fire deecion sysem is required. In case of 3 dimensions model he balance beween he mehods of aerial fligh parol or he ower based fire deecion sysem depends on feaure and rae of he area ariculaion. The 4 dimensions analysis says ha he balance beween he mehods of aerial fligh parol or he ower based fire deecion sysem can correlae o he fire weaher index. Keywords: aerial parol, UAS applicaion, ower based fire deecion sysem, dimension analysis. Inroducion One of he key elemens of he cheap bu effecive fire suppression is early fire deecion and quick iniial response. There are many ways of soluions for he fire deecion, such as saellie in orbial equipped wih special sensors, manned or unmanned aerial (UAV) fligh parol a differen aliude, many kinds of ower based auomaic fire deecion sysems, or mobile human surveillance. Each of hem has operaion coss, in many cases no negligible, herefore aking an overview of he effeciveness is no useless. Saellie can be a very effecive ool for fire deecion bu here are some problems wih he applicaion. Saellies a geosaionary can monior huge areas bu he disance is oo far from he Earh o deec ho spos in small size. Sensors wih higher resoluion could help o deec smaller size fire bu here are oher problems, like higher price or more false alarms (Pennypacker, 03). Saellie wih closer orbi han geosaionary can be also a nice soluion bu keeping he disance from he Earh requires higher speed meaning ha he same area can be moniored cyclically (Giglio a al., 008). In his case we can coun a moniored ime and a blind ime; laes means ha area is no moniored coninuously (Resas, 0). This saellie can be a good soluion a huge foresed areas such as Norh and Souh America, Africa or Siberia. Human surveillance is also a kind of fire deecion mehod, however he effeciveness is many cases accidenal. All mehod menioned above belongs o sraegic soluions for ho spo deecion however a he beginning of he inervenion managers also require acical soluions. I can mean horizonal observaion from he ground or human wih special echnical equipmen like fire ruck or UAS.. Aerial parol for ho spo deecion dimensions analysis.. Basic elemens of he effeciveness The basic assumpion is ha by using aerial parol he fire service can save fores of more value han wihou i. The economic efficiency will maerialize if he saved fores is more han he all expendiure of fire service 76
regarding using aerial surveillance. A he sraegical level, auhor means a bigger scale, le s say a a governmen level, we have o ake ino accoun all he expendiure of aerial surveillance and all he saved fores of he counry. Aerial parol can deec ho spos very quickly and i is able o give he firs fire repor o fire brigades; i can reduce he ime of firs aack. Figure Aerial surveillance e.g. using UAV for ho spo deecion For ho spo deecion spoer makes a parol following he pre-programmed fligh pah usually as rouine ask. Obviously, he average delay of he aerial fire repor mus be less han he average delay of he civil repor. This condiion is necessary even if no sufficien for saisfying he requiremens of economic efficiency of aerial surveillance, in his example UAV fligh (figure ). From he igniion (fire sars) ill he aerial based fire repor akes differen ime depending on he posiion of he UAV and he posiion of he ignied fire crossed by planned fligh pah. Obviously large number of fire deecion (saisical daa base) gives for he ime of average fire repor as a half ime of he UAV parol ime. This saemen can be acceped also by logical conclusion. UAV _ repor UAV _ parol () UAV _ repor Civil _ repor () - UAV_parol - full ime of a fligh parol, made by UAV; - UAV_repor - average ime of fire repor given by UAV, making fire parol; - Civil_repor - average ime of fire repor, given by civilians. The oher crieria for saisfying he economically efficiency is ha based on aerial surveillance (e.g. UAV use) fire service mus save fores of more value han he coss of he surveillance (aerial parol). A quick fire repor resuls in quick answer; shor unconrolled fire means less damages bu more saved value. If he inervenion sars very soon afer igniion, he savings will come no jus from he saved fores bu also from a shorer use of he special equipmen required o suppress he fire. Figure Srucure of damage ime funcions aking ino accoun he quick fire repor Modelling he exinguishing of he fire by he damage ime funcions he beginning of he response will break he fire curve (figure ). The inensiy and he lengh of he modified fire curve will depend on he ime beween he igniion and he beginning of he response. In case of A he ime is shorer han in case of B herefore in case of A he inensiy is lower and he lengh is shorer of he modified fire curve han in case of B (figure ). The 77
response in case of A resuls in lower damage (los) and shorer ime of inervenion; his laes one resuls also in lower coss of he inervenion. The shorer he ime is beween he igniion and he beginning of he response, he lower he inensiy and he shorer he modified fire curve is. The above crieria can be expressed also by he nex formulas: S D Civil _ repor UAV _ parol C UAV _ parol (3) Civil _ repor D UAV _ parol C UAV _ parol (4) - S (D) - damages beween he average erm of civil repor and aerial repor; - A - characerisic of suppression using UAV suppor; - B - characerisic of suppression wihou UAV suppor. Till he above formulas are valid he crierion of he economically based efficiency is saisfied a UAV parol based fire managemen... Moving o higher effeciveness fligh speed analysis Since he efficiency obviously depends on he difference of he civil based and aerial supplied repor s ime, he quesion sponaneously arises, how can we make longer erm beween hese repors? Based on he above formulas he delay of civil repor is relaively sable, while he aerial based ho spo deecion depends mainly on fligh regimes. Therefore, making he aerial based ho spo deecion more efficien we mus face fly parameers. Logically we can es he fligh speed, he aliude and he visual or camera (UAV) focus. During he analysis idealisic circumsances are supposed: here is no wind, i couns jus wih one ho spo, price of he echnical elemens, like camera does no change, weaher does no limi he visibiliy, area is fla, ec. This laes one, he fla area means ha we monior dimensions exension. The purpose of reducing he average deecion ime requires he fligh speed rise. Wih his process he fligh pah will no change bu he ime of aerial parol will reduce. In case of a 4 km x 4 km area, wih abou a 80 kmh - fligh speed and abou a 3 km x 3 km moniored (camera) pixel in a momen (e.g. UAV) he moniored ime for pixel is minue hen 63 minues are blind; he rae of moniored and blind area is :63. Raising he speed he cycle will reduce bu he rae of he moniored and blind area will no. The problem is ha he fligh speed rise is objecively limied (figure 3). Figure 3 Example for moving owards he higher effeciveness of he fire deecion. Raising he fligh speed neiher he fligh pah nor he rae of moniored and blind area do no change.3. Moving o higher effeciveness camera focus and aliude analysis 78
The analyses of camera focus and aliude gives a lile bi differen resuls. Rising he aliude he on board insalled camera moniors larger area (pixel). If he frame of he supervised area remains bu he moniored pixel is larger han earlier, i means ha he fligh pah can change; obviously i will be shorer. The fligh pah will move o he cener of he area. Sep by sep coninuing his process i reaches he end values when he fligh pah will concenrae a he middle poin of he supervised area. I means our aerial means hangs on he cenral poin of he area wihou fligh speed! Logically he end value doesn require he manned or unmanned fligh; in his case a ower based camera sysem can be an alernaive soluion wih he same effeciveness. Figure 4 Example for moving owards he higher effeciveness of he fire deecion. Raising he angle of camera focus or he aliude he fligh pah changes; i moves o he cenre of he moniored area The analysis of camera focus wih similar way as above gives he same resuls (figure 4). In his case here is no or insignifican difference beween he aliude of aerial surveillance hanging on he cener poin of he area and ower based fire deecion sysem. In boh cases he rae of he moniored and blind area also changes; i drasically urns beneficially. A he end of he process he rae is :0; which means all area is moniored coninuously. Even if he effeciveness of boh sysems aerial surveillance and ower based fire deecion sysem can be similar ha is no sure in case of efficiency. If all he coss of using ower based fire deecion sysem and he aerial surveillance are similar he efficiencies are also similar. In he assumpions we ook ino accoun same echnical parameers i means camera insalled on board or ower oo herefore, he difference of he efficiency generaes from he all coss of using he sysems. 3. Geographically high ariculaed area 3 dimensions analysis Based on he above using ower based fire deecion sysem in case of fla area dimensions exension is more effecive. The quesion is ha in wha or when he higher efficiency can be limied. As a firs sep, auhor exended fla dimensions area o geographically high ariculaed 3 dimensions area. In case of fla area ower based fire deecion sysem can deec ho spos by direc way; i means here is no naural or manmade barrier o see he ho spos by he camera or sensor. In case of ariculaed areas he way for fire deecion can be differen. If he igniion is a he valley or a he boom of he slope i can be in shadow of he hills, herefore, ower based fire deecion sysem can deec fire in many cases only by indirec way; i means fire is in he shadow of he hill, camera or sensor canno deec he fire bu he smoke column or plum of i (figure 5). Naurally he smoke column in indirec way can be deeced laer han fire in direc way. Depending on differen circumsances wind power and direcion, inversion, fire inensiy, rae and way of he ariculaion, ec. he deecion of smoke column can delay significanly. I means ha no jus he effeciveness of he ower based fire deecion sysem reduces bu also he efficiency of i. The rae of reducion naurally depends on he rae and way of ariculaion; his is a circumsance ha does no change in ime. In case of high ariculaion reducion of 79
effeciveness is surely higher; however, he rae of correlaion beween hem requires advanced research. Moreover, he way of ariculaion is also imporan. Posiioning his kind of sysem is very imporan; i basically deermines is poenial of boh effeciveness and efficiency. Good posiioning even if jus a limied way, bu can reduce he shadow effec of hills. When posiioning of he sysem canno be advanageous bu he area is high ariculaed he rae of direc deecion and indirec deecion can be also very bad. The smoke emission from he valley sand ignied fire can be deeced only wih such delay so he advanages of low yearly coss evaporae. Obviously he mission coss of aerial surveillance do no depend or jus o a limied exen on he rae of area ariculaion. On he oher hand, during fligh parol he shadow effec of hills can be relevan only for ha pixel where he aircraf flies above. Naurally his effec is much lower in his case han in case of ower based fire deecion sysem. Since he effeciveness and efficiency correlae o each oher, he higher coss of aerial surveillance can be balanced by he lower effeciveness of he ower based sysem. I can be explained by oher words oo; he rae of caching ho spos in ime by aerial surveillance is higher han in case of ower based sysem. The disadvanageous rae of moniored and blind area is balanced by higher rae of quick fire repor regarding shadow area. The above crierion can be expressed by formula () and (5) oo. Figure 5 High ariculaed area: direc deecion versus indirec deecion (red colored slopes). Aerial surveillance versus ower based fire deecion sysem UAV _ repor UAV _ parol () UAV _ repor delay _ ower _ repor (5) - delay_ower_repor -average delay of fire repor, given by ower based fire deecion sysem. As he formulas () and (3) are logically same a ime facor, formula (3) and (6) also have logical similariy in coss facor. S D delay _ ower _ repor UAV _ parol C UAV _ parol (6) I means also ha i mus saisfy oher basic crierion oo, as quicker average fire repor han repor of civilians. Till he above formulas are valid he crierion of he economically based efficiency is saisfied a aerial surveillance versus ower based fire deecion sysem. 4. Exremely High Fire Weaher Index 4 dimensions analysis Analysis above focused on physical exension of he given area; dimensions couned as fla area, 3 dimensions as ariculaed area. In any case, here is no difference in ime range; i akes ino accoun ime range as sandard condiion, however areas are hreaened by fire risk usually cyclically so-called fire seasons. If he fire risk is cyclic, he quesion arises sponaneously, can i be even cyclic or no monioring he area? If we ake ino accoun he cyclic fire seasons as ime facor, i can mean as 4 dimensions analysis. I is known, fire seasons srongly correlae o weaher condiions; for measuring i differen ype of index fire weaher index (FWI) were creaed, e.g. Canada: McArhur FWI (Dowdy a al., 00), Germany: Waldbrandgefahrenindex WBI (König, 007). Tha is obvious, area monioring is he more required he higher 80
he fire risk; in his case probabiliy of maching ho spos is higher, which is shown by FWI. Therefore, i is pracical o find correlaion beween FWI and effeciveness of fire deecion (figure 6). Figure 6 Examples for Waldbrandgefahrenindex, July 04, Germany Following he correlaion, FWI based planning of applied sysem for fire deecion does no require area monioring all year round. This condiion is no relevan for ower based fire deecion sysem because is insallaion fix and i is able o work during all year. I means we need coun wih is all year coss. In case of aerial surveillance we have o coun wih coss only for ha period when i is occasionally demanded. There are some opions for analyzing he efficiency: areas exension can be aken as dimensions or 3 dimensions, however he essence of boh resuls mus be similar. In case of same effeciveness he average rae of fire deecion is equal he cheaper soluion akes he higher efficiency, however his possibiliy has low chance. The oher siuaion is when here are differences beween boh he average raes of he fire deecion and he coss of using he differen sysems; obviously he higher coss of he aerial fire deecion are suspeced. In case of same efficiency he higher coss of aerial fire deecion mus be balanced by he higher rae of maching ho spos; i is he ipping poin. Above his hreshold aking ino accoun fix price for he fligh hour aerial fire deecion is more efficien, oherwise no. Oherwise, if here is difference in mach of ho spo deecion we have o calculae he rae of he difference beween he saved fores and coss. In basic siuaion he crierion can be expressed by he nex formula: C UAV _ occasionally C Tower _ yearly _ cos s (7) C UAV_occasionally - occasional coss of UAV, used i a exremely high FWI; C Tower_yearly_coss - yearly coss of he ower based fire deecion sysem. Till he above formula is valid he crierion of he economically based efficiency is saisfied a exremely high FWI erm. I can be seen, his laes analysis conains he mos vulnerable elemens. Therefore, his paper jus shows some direcions for furher research, exac definiion of he ipping poin as well as deermines he correlaion beween he higher rae of maching ho spos and he efficiency requiring more sudy. 5. Summarizing This paper firsly described he basic principles of he economical effeciveness of he fire managemen. Aferwards, research urned o he differen dimensions and basically couns wih aerial fligh parol. Looking a he moniored fores as a fla area, i gives dimensions analysis bu as an ariculaed area i gives 3 dimensions analysis. Research couned also wih he ime scale, meaning ha weaher condiions can change; i gave he 4 dimensions analysis. Based on he research in he dimension model he aerial fligh parol canno be effecive, a ower based fire deecion sysem is required. In case of 3 dimensions model he balance beween he mehods of aerial fligh parol or he ower based fire deecion sysem depends on feaure and rae of he area ariculaion. The 4 dimensions analysis says ha he balance beween he mehods of aerial fligh parol or he ower based 8
fire deecion sysem can correlae o he fire weaher index. Research poined ou ha applied analysis conains many vulnerable elemens. Therefore, his paper jus shows some direcions for furher research, exac definiion of he ipping poin as well as deermines he correlaion beween he higher rae of maching ho spos and he efficiency requiring more sudy. References Dowdy, A.D., Graham A. Mills G.A., Finkele, K., Groo, W.: Index sensiiviy analysis applied o he Canadian Fores Fire Weaher Index and he McArhur Fores Fire Danger Index; Meeorological Applicaions, 7 (3), pp. 98 3, Sepember 00 DOI: 0.00/me.70 Giglio, L., Csiszar, I., Resas, A., Morisee, J.T., Schroeder, W., Moron, D., Jusice, C.O.: Acive fire deecion and Characerizaion wih he Advanced Spaceborne Thermal Emission and Reflecion Radiomeer (ASTER) Elsevier Science, Remoe Sensing of Environmen (008) 3055-3063 König, H.C.: Waldbrandschuz. Kompendium für Fors und Feuerwehr. Ediion GefahrenAbwehr; Supplemen (Band ). Fachverlag Grimm, Berlin 007, 97 S., ISBN 978-3-94086-0-7 oder ISBN 3-94086-0-X Pennypacker, C.R., Jakubowski, M.K., Kelly, M., Lampon, M., Schmid, C., Sephens, S., and Tripp, R.: FUEGO Fire Urgency Esimaor in Geosynchronous Orbi A Proposed Early-Warning Fire Deecion Sysem; Remoe Sensing, 03 5(0) ISSN 07-49 Resas, A.: An Approach for Measuring he Economic Efficiency of UAV Applicaions a Fores Fires Helping Decision Makers; AUVSI 0: The Firs Inernaional Conference on Unmanned Sysems in Israel. Tel Aviv, Israel, 0. CONTACT ADRESS Auhor: Workplace: E-mail: Agoson Resas, PhD Naional Universiy of Public Service, Budapes, Hungary Resas.Agoson@uni-nke.hu RECENZIA TEXTOV V ZBORNÍKU Recenzované dvomi recenzenmi, členmi vedeckej rady konferencie. Za exovú a jazykovú úpravu príspevku zodpovedajú auori. REVIEW TEXT IN THE CONFERENCE PROCEEDINGS Conribuions published in proceedings were reviewed by wo members of scienific commiee of he conference. For ex ediing and linguisic conribuion corresponding auhors. 8