Selecting boilers in an energy flexible heating systeme based on lowest running cost

Transcription

1 10th International Symposium on District Heating an ooling September 3-5, 2006 Monay, 4 September 2006 Sektion 4 b onceptions, rafts an stuies in istrict heating an cooling Selecting boilers in an energy flexible heating systeme base on lowest running cost R. Riise, B. Jensen, Narvik University ollege/norway

2 Selecting boilers in an energy flexible heating system base on lowest running cost M.Sc, Riise, Raymon, M.Sc, Jensen, Bjørnulf, PhD, Sørensen, Bjørn Reiar, Narvik University ollege, Department of Builing, Prouction an Engineering Design, P.O. Box 385, N-8505 Narvik, Norway. Phone: Abstract This paper presents two ifferent methos for calculating the final energy requirement from the heating system per time step one with a preefine yearly mean efficiency an one with calculate efficiency for each time step. Further the paper presents a simple metho for selecting the most cost efficient boiler, base on lowest running cost, an the possible consequences of selecting wrong boiler. A simulation example emonstrates the effect of selecting boilers base on yearly mean efficiencies rather than calculate mean efficiencies for the actual time step in question. Keywors: Selecting boilers; Yearly mean efficiency; Actual efficiency; ontrol strategy; Lowest running cost 1. Introuction Selecting the wrong boiler for heat prouction can result in unnecessary use of energy an may be expensive if the conition is allowe to continue. In many cases the operations manager o not know how to select the most cost efficient boiler for each time step base on varying outoor temperature, loa pattern an energy prices. Often use as criteria is the yearly mean boiler efficiency. These preefine efficiencies are often introuce to the operations manager by people how o not know the actual heat station in question an therefore can not know what exact efficiency to use. Very often operators are tol to use 75% efficiency on oil fire boilers an 95% on electrical boilers. This paper presents two ifferent methos for calculating the final energy requirement from the heating system per time step one with a preefine yearly mean efficiency an one with calculate efficiency for each time step. Further the paper presents a simple metho for selecting the most cost efficient boiler, base on lowest running cost, an the possible consequences of selecting wrong boiler. The process is ivie into three steps. First the total energy requirement to the heating system is calculate (as escribe in section 2). Then the expecte final energy requirement is calculate base on the two ifferent methos of boiler efficiencies (as escribe in section 3), an finally the selection of the most cost efficient boiler base on lowest running cost (as escribe in section 4). 2. Total energy requirement to the heating system First the total energy requirement by the builing to the heating system must be calculate.

3 The total energy requirement to the heating system,, consists of energy requirement for space heating (incluing infiltration), mechanical ventilation an omestic hot water + + [kwh] (2.1) h v w h v w energy requirement for space heating [kwh] energy requirement for ventilation [kwh] energy requirement for omestic hot water [kwh] Heat eman for space heating, h, without taking into account the system losses are calculate uner stanarise conitions, accoring to EN 832, EN ISO or similar. In this paper a simplifie metho for calculation of heat loss through the builing envelope, h, has been use h H ( θ θ ) t [kwh] (2.2) in e H total heat loss coefficient of the builing [W/K] θ in inoor set-point temperature [ o ] θ e average external temperature uring the calculation perio [ o ] t uration of the calculation perio Heat eman for ventilation, v, is calculate by v ( η )( θ )t c ρ V 1 θ [kwh] (2.3) v in e c air specific heat capacity [kj/kgk] ρ ensity or air [kg/m 3 ] V airflow rate [m 3 /s] η v heat recovery efficiency [ ] θ in supply air set-point temperature [ o ] θ e average external temperature uring the calculation perio [ o ] t uration of the calculation perio Heat eman for omestic hot water, w, is given by w ( θ ) t ρ c V θ [kwh] (2.4) w w o ρ ensity of water [kg/m 3 ] c specific heat capacity of water [kj/kgk]

4 V w volume of hot water require [m 3 /s] θ w temperature of the hot water [ o ] θ o temperature of water entering the omestic hot water system [ o ] t uration of the calculation perio In this paper the heat eman for omestic water is set for an example builing, see the simulation example in section Final energy requirement by the heating system Next the final energy requirement by the heating system must be calculate. The correct selection of boiler base on lowest running cost for one time step, for instance one hour, necessitates calculation of the require final energy (net energy) for that exact time step. Final energy requirement, f, can easily be calculate when the heat eman neee to be elivere from the heating system an the efficiencies of each boiler is known f η [kwh] (3.1) l : total energy requirement to the heating system [kwh] η l efficiency of boiler l [ ] The next two subsections present two ifferent methos for calculation of final energy requirement base on two ifferent approaches of efining the boiler efficiencies. 3.1 Final energy requirement with preefine fixe yearly mean efficiency The first metho for calculating final energy requirement is base on yearly mean boiler efficiencies. A preefine fixe yearly mean efficiency for each boiler is use to calculate the expecte final energy, ef, base on the total energy requirement. ef [kwh] (3.2) η l, m : total energy requirement to the heating system [kwh] η l,m assume yearly mean efficiency for boiler l [ ] This approach is wiely use real efficiencies are not known. Often an efficiency of 75% is use on oil fire boilers an 95% for electric boilers. In some cases operators even use the boiler efficiency with flue gas loss as the only heat loss from the oil fire boiler which is given as the boilers efficiency by service personnel.

5 When using this metho it is most likely that the calculate expecte final energy requirement for each time step will iffer compare with actual energy consumption, an hence there is a possibility that the most expensive boiler alternative will be chosen. 3.2 Final energy requirement with mean power efficiency for each time step The secon metho for calculating final energy requirement is base on mean boiler efficiencies for each time step calculate as a function of the heat eman from the builing for the same time step. Nielsen /1/ presents a set of equations to fin the efficiency as a function of the heat eman from the heat system for an intermittently controlle boiler. These equations have been moifie to take into account relative istribution heat loss, L. Loss in the istribution system will influence on the intermittency of the boilers an hence the efficiency. out η l ( Il + 1) (3.3) inst The intermittence rate can be obtaine by using equation (3.4)-(3.7) as follows: Nl 1 I l (3.4) Nl + 1 I l,max N l inst η out l,max max out (3.5) I I o,max e,max 1 Leg Lr L L + L + L r r g 1 Lr L L L (3.6) η η o,max e,max 1 L 1 L eg r L L r L (3.7) η l mean efficiency for the boiler in question [ ] out heat output from heat system [kwh/h] inst gross installe boiler capacity [kwh/h] max maximum heat output from heat system I intermittence rate [ ] N help variable I max maximum intermittence rate, when η0 [ ] η max maximum operating efficiency (I0) [ ]

6 L eg relative exhaust gas loss [ ] L r relative raiation an convection heat loss for boiler surface [ ] L relative istribution heat loss [ ] L g relative heat loss cause by air raught at boiler off time [ ] Losses from each boiler must be obtaine from calculations, measurements or from the boiler manufacturer. When the efficiency for each boiler for the time step in question have been calculate it is a simple matter to etermine the final energy requirement, f, by combining equation (3.1) an (3.3) for that time step: f η l out inst ( I +1) l [kwh] (3.8) This approach is not wiely use in cases the selection of boilers is one manually, but it have become more an more common in automate heat stations a control unit selects the optimal boiler base on lowest running cost. 4. Selecting boiler base on lowest running cost Finally selection of the most cost efficient boilers can be aresse. When the final energy requirement for the time step is calculate the selection of boilers can be one base on lowest running cost. For this, the energy purchase price per kwh must be known. For an electric fire boiler the total running cost, e, for each time step is: e c [money/h] (4.1) f e f c e final energy requirement by the heating system [kwh/h] purchase price for electricity [money/kwh] For an oil fire boiler the total running cost, o, is: o c [money/h] (4.2) f o k o c o [money/kwh] (4.3) hn f c o k o final energy requirement by the heating system [kwh/h] purchase price for oil [money/kwh] purchase prise for oil [money/litre]

7 h n lower calorific value per litre oil [kwh/litre] It is now a simple matter to calculate which boiler to choose: o F (4.4) e F the rate between the efficiencies [ ] If F > 1 choose electric boiler F < 1 choose oil fire boiler F 1 choose from preferre criteria other than running cost, e.g. environmental impact. 5. onsequences from selecting wrong boilers When etermining final energy requirement by using yearly mean efficiency it is most likely that calculate expecte final energy requirement will iffer from the real energy consumption since the real efficiency for the time step in question most likely will iffer from the assume yearly mean efficiency. alculation using the mean boiler efficiencies for each time step will be more accurate since these efficiencies are calculate as a function of the heat output from the boiler at the same time step. When etermining final energy requirement per time step base on yearly mean efficiencies only the expecte running cost is calculate, not the real running cost. The real running cost can only be calculate with use of the exact efficiencies per time step. In this paper the time step is set to one hour since external temperature an electricity purchase costs are given on an hourly basis. To be able to emonstrate the effect of selecting the wrong boiler both methos presente in the paper must be use to calculate running cost per time step. When selecting boilers base on lowest running cost from calculations with yearly mean efficiencies an the actual energy price the wrong boiler can be chosen. Table 1 shows a simple example of a typical situation. Table 1 Example of calculations alculation metho,i [kwh/h] η f,i [kwh/h] Energy price Time step i [NOK/kWh] Total energy cost time step i [NOK] Oil El Oil El c o,i c e,i o,i e,i Yearly mean efficiency η m Hourly mean efficiency η i The rate between the yearly mean efficiencies is: F o, i e, i

8 Since F > 1 the electric boiler is chosen accoring to the efinition in section 4. The rate between hourly calculate mean efficiencies is: F o, i e, i Since F < 1 the oil fire boiler is chosen accoring to the efinition in section 4. In this example, when the choice of boiler was base on yearly mean efficiencies, the electric boiler woul have been operate for time step i. alculations with hourly mean efficiencies show that the correct choice woul have been the oil fire boiler. By running the electric boiler the total running cost woul have been NOK, but the total cost coul have been only NOK if the oil boiler chosen. The calculate expecte running cost of NOK an NOK are only expecte cost calculate with an assume efficiency. The real running cost are NOK for the oil fire boiler an NOK for the electric boiler. 6. Simulation example Base on previous sections a simulation example that emonstrates the effect of running a heat station for a year base on inaccurate boiler efficiencies is given. 6.1 Builing escription Total energy requirement to the heating system have been calculate accoring to section 1 with the builing ata shown in table 2. The temperature istribution for each 24-hour perio are estimate base on formula given by Programbyggerne /3/ + ˆ t tmax Te ( t) Te Te cos π [ o ] (6.1) 12 T e (t) external temperature at time t [ o ] T mean external temperature uring 24-hour perio [ o ] e Tˆ e temperature amplitue uring perio [ o ] t hour of ay [ ] t max hour of ay when maximum temperature occurs (here t max 13) [ ] Average external temperatures are shown in figure 6.1. Only one boiler is allowe to run at each time step, an both boilers net capacity are larger than the heat eman uner esign climate conitions.

9 Table 2 Data for the simulation example Type of builing Office builing Year of construction 1954 Inoor set-point temp 17 o Area 3100 m 2 Volume 9900 m 3 Infiltration rate 0.3 h -1 Total heat transfer incl. infiltration 3258 W/K Heat eman omestic hot water 5 W/m 2 Air specific heat capacity 0.35 Wh/(m 3 K) Airflow rate m 3 /h Heat recovery efficiency 0.75 Operating time ventilation (Monay-Sunay) Oil boiler gross installe capacity 400 kw Electric boiler gross installe capacity 350 kw External temperature Daily maximum an minimum temperature given by Norwegian Meteorological Institute. Location Narvik. Year 2005 Electricity prices Hourly electricity prices given by Nor Pool ASA. Figure 6.2 shows average aily prices for 2005 incluing transmission charges, NOK/kWh, from local power gri owner, Narvik Energi AS Oil prices Approximately marke price for fuel oil for For simulation purposes it is assume ifferent prices for each month of the year. Figure 6.3 shows prices for each month use in the simulation example. Lower calorific value 10 kwh/litre Specific oil boiler losses L eg 8.5% L r 3.5% L g 2.0% L 0.6% Specific electric boiler losses L r 2.5% L 0.6 % Average external temperature Degrees elcius Time Figure 6.1 Average external temperature Narvik 2005.

10 Average aily electricity prices NOK/kWh Time Figure 6.2 Average aily electricity prices incluing transmission cost. Oil prices NOK/litre Time Figure 6.3 Average monthly oil prices. 6.2 Results of simulation Final energy requirement With yearly mean efficiency With hourly calculate mean efficiency Total running cost With yearly mean efficiency With hourly calculate mean efficiency kwh kwh NOK NOK Number of hours running the most expensive boiler By controlling the boilers base on yearly mean efficiencies an actual energy prices, the most cost expensive boiler was operate hours resulting in an unnecessary energy consumption of kwh at a cost of NOK. Figure 6.4 isplays the various efficiencies etermine in the simulation an the preefine yearly mean efficiencies for both boilers.

11 Average efficiency per ay Efficiency Time Real hourly mean, oil boiler Real hourly mean, electric boiler Assume yearly mean, oil boiler Assume yearly mean, elcetric boiler Figure 6.4 Average boiler efficiency per ay. Figure 6.5 shows the average running cost per ay for both control strategies. Average running cost per ay NOK/h See figure Time alculate with yearly mean efficiency Figure 6.5 Average running cost per ay. alculate with hourly mean efficiency A closer inspection of the curves of figure 6.5 is shown in figures 6.6 an 6.7.

12 Runnig cost per hour week oncurrent curves inicates operating the most cost efficient boiler Length of eviation inicates number of hours operating the wrong boiler NOK/h Deviation inicates operating the wrong boiler. Size of eviation inicates size of unnecessary energy cost Time alculate with yearly mean efficiency Figure 6.6 Running cost per hour for week 28. alculate with hourly mean efficiency Figure 6.6 shows the eviation in the control strategy with yearly mean efficiencies compare with hourly mean efficiencies occurs. Where there cost curves are concurrent the most cost effective boiler was operate even though the ecision basis was wrong. Where the curves eviate the wrong boiler was operate. Figure 6.6 only shows the cost eviation in the two control strategies. Figure 6.7 shows which boiler who was operate at each time step in week 28. Selecte boiler Oil fire boiler Selecte boiler Electric boiler Yearly mean efficiency control strategy Time Figure 6.7 Selecte boilers uring week 28. Hourly mean efficiency control strategy The curve inicating the control strategy using hourly mean efficiency shows that the correct choice was the electric boiler for the whole perio, but the yearly mean efficiency control strategy selecte the more expensive oil fire boiler for 102 hours for week Other influences on boiler selection It is ifficult to be specific about possible cost savings by always selecting the boiler with lowest running cost for each time step. There are many ifferent factors that can affect the final result.

13 In this paper start-up an shutown costs for each boiler are not taken into account, only running cost. Start-up an shutown costs will occur every time a boiler is starte or shut own ue to preheating an cooling of the boiler. Nilsen /3/ states that smaller boilers may have negligible shutown costs but the start-up cost shoul be consiere. Energy consume uring start-up perio epens on the length of the perio without prouction ue to cooling of the unit. In aition to varying energy costs there will also be fixe costs such as labour hours an extra costs ue to wear associate with start-up an shutown of boilers. Figure 7.1 shows an example of how start-up costs vary as a function of time without prouction. Start-up costs Start-up costs [NOK] Time without prouction [hours] Running start-up cost Fixe start-up cost ol start-up Figure 7.1 Example of varying start-up costs In the simulation in section 6 the total number of start an stops are 455 for calculation with preefine yearly mean efficiency an 460 for calculation with hourly mean efficiency. The major problem etermining the start-up cost is that it is ifficult to preict for how long the selecte boiler will be operate until it stops again. This makes it ifficult to istribute the start-up cost for each time step. When start an stop costs are introuce in a moel this cost must be taken into account on top of calculate running cost when choosing which boiler to use. Introuction of start- an shutown cost for each boiler may result in fewer switches between boilers if the margins in running cost are small. In aition to the purchase cost for electricity, money/kwh, often consumers have to pay a transmission charge for power, money/kw, base on the size of maximum use power tap uring the year. These charges can be significant. The problem with aing these transmission charges is that we can not state until the last hour of the year how large this cost will be. In some cases a maximum power restriction can be set to the heat station to limit the transmission charge per kw. More accurate climate moel incluing a more accurate temperature preiction win an solar raiation will improve the accuracy of the calculate heat eman an hence improving the accuracy of the final heat eman from the heating system. This will give a more accurate result. 8. Other possible selection parameters In some cases builing owners may have other parameters than lowest cost for optimizing the heat prouction. Such criteria may by: - lowest possible energy consumption

14 - lowest possible emission of air pollution - highest possible use of renewable energy. For most cases lowest possible running cost will be the optimization criteria of choice, but for some builings other parameters, such as mentione above, may be ecisive base on governmental or private policies. In such cases the selection moels must be ajuste to each iniviual selection criteria before choosing which boiler to use. 9. onclusion This paper has presente two ifferent methos for calculating final energy requirement from the heating system per time step - one with a preefine yearly mean efficiency an one with calculate efficiency for each time step. Further the paper presente a simple metho for selecting the most cost efficient boiler, base on lowest operating cost, an the possible consequences of selecting wrong boiler. A simulation example reveale that it is important to use actual efficiencies, rather than preefine yearly mean efficiencies, for each boiler when calculating the running cost base on actual heat eman an actual energy prices. Selecting the wrong boiler can be expensive. References /1/ Nielsen, John Rune (1996): A moel for optimization an analysis for energy flexible boiler plants for builing heating purposes. Thesis for the Degree of Dr. ing., Norwegian Institute of Technology. /2/ Programbyggerne (2002): User manual Energi i Bygninger 3.5. Programbyggerne. /3/ Nilsen, Jan Øyvin (1994): A computer moel for planning of energy system with time epenent components an bounary conitions. Thesis for the Degree of Dr. ing., Norwegian Institute of Technology.