- Email: [email protected]
Implementation of a cogenerative district heating: Optimization of a simulation model for the thermal power demand
Energy and Buildings 38 (2006) 1434–1442 www.elsevier.com/locate/enbuild
Implementation of a cogenerative district heating: Optimization of a simulation model for the thermal power demand L. Barelli *, G. Bidini, E.M. Pinchi Dipartimento di Ingegneria Industriale, Universita` degli Studi di Perugia, Via G. Duranti, 67-06125 Perugia, Italy Received 12 November 2005; received in revised form 17 March 2006; accepted 20 March 2006
Abstract The district heating set up with a cogeneration system, concurs to attain energetic, economic and ambient benefits. It also provides to citizens a new service. The project strategy is based on the idea of supplying a portion of the necessary thermal power through a combustion alternative engine in cogeneration modality. It’s also interesting to modulate the load with auxiliary boilers fed by natural gas. This solution allows to save primary energy, create a centralization of the energy production, which contributes to the problem of polluting emissions, through the decentralization of the sources. The first step to assess the technical-economic feasibility of a district heating system, based on a cogeneration plant, is to underline and to characterize the energetic request of the basin of user. The objective of the present work is to develop a model that yields an esteem of the hourly thermal load for every days of the heating season of a complex user, represented by a single neighbourhood. To do this, the present work proposes a new method of simulation of the daily and hourly thermal load trend, known only the value of the power installed in the thermal plant for every user, the seasonal hours of the burner operation and the timetable of the heating service distribution, more than the external mean daily temperature trend. The results obtained using this model, have been verified with the data of seasonal consumptions, confirming the validity of the proposed methodology.The above allows to determine, with more precision, the thermal request peak to satisfy, taking in consideration the contemporaneity of the loads, also of different typology, and to carry out a better sizing of the generation plant. # 2006 Elsevier B.V. All rights reserved. Keywords: District heating; Simulation of the thermal demand; Cogeneration
1. Introduction At this moment the situation in terms of energy and energetic resource, results critical. The stock of the fossil fuel decreases gradually and the increase of the international energy request has produced a search of objectives, ratified in the Kyoto protocol, such as the rational use of energy, the improvement of the process’s efficiency, the passage towards fuel with a minor contained of carbon. But, the principal objective is to reduce the atmospheric emissions which cause the greenhouse effect. The present work proposes to develop a model of the thermal daily and hourly load of a complex user such as a neighbourhood during the heating season. This to value the thermal power peak necessary to correctly size the generation plant supplying heat to the district heating system [1–8]. The
* Corresponding author. Tel.: +39 075 5853740; fax: +39 075 5853736. E-mail address: [email protected] (L. Barelli). 0378-7788/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2006.03.023
model gives the seasonal amount of the thermal energy necessary to the buildings of the basin of user, to guarantee to the users, in function of the trend of the external mean seasonal temperature, the comfort requested. Knowing the total request, pointing out the better configuration of the operation, it’s possible to give a realistic evaluation of the fuel consumption necessary, the corresponding level of the emission in the environment, the electrical energy production and the energetic saving attainable. The strategy chosen is foreseen to guarantee, with a combustion alternative engine in cogeneration modality, a thermal power satisfying the thermal request characterizing the user for all of the heating season, supplying the remaining power, when necessary, through traditional boilers. In particular the study relates to the buildings, of an area comprising some neighbourhoods in Perugia, a city in central Italy, in order to arrange a district heating system combined with a cogeneration plant. This to increase the saving of primary energy obtainable, to realize the centralization of the
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
Nomenclature C DD DH Ti Tme w
global heat loss Degree Day Degree Hour internal temperature mean daily external temperature weights distribution
energy generation and the delocalization of the polluting production, controlling at the same time the emissions using suitable control systems. As a first step a census of the user of this study case has been conduced; the majority of the population is composed by residential buildings and some schools. The collected data for every thermal plant of the user, concern the combustible typology used, the seasonal consumptions, the power currently installed, the ignition hours of the burner and those of the heating service. For every user typology, a reference model deduced from experimental data has been conceived, in order to characterize the request of the typical user. This involves the determination of the weights distribution in 24 h defined as the ratio between the thermal mean hourly power distributed (valued as the mean of the thermal hourly powers supplied in the days of the surveys) and the thermal mean power distributed, calculated starting from the available data of all the surveys realized in the hourly ignition of the heating system. Having determined such weights distribution, it’s possible to assess the hourly trend of the inner temperature truly guaranteed with the current thermal plant. Then the weights distribution were varied through suitable modifications in the heating system operation, implying the reduction of the thermal power peaks requested in the thermal transitory at the ignition system. The subsequent analysis of the hourly trend of the inner temperature, allows to verify the satisfaction of the thermal comfort requested. To optimize this procedure and to obtain a more uniform daily trend of the inner temperature, the thermal transitory, considered as the period necessary to have the regime conditions in the building, was investigated. A modified thermal profile of the request, which characterizes the typical user, is obtained. This methodology was subsequently extended to the entire heating season, utilizing the weights distribution previously obtained and creating the correlation between the external temperature trend and the thermal request. Calculation procedure was finally extended to all the typologies of users, obtaining the daily and hourly trend of the total thermal load. On this basis the methodology was verified, through the true consumption data and the generation efficiency (valued starting from the hours available data), showing the efficiency of the model proposed.
1435
residential buildings. For all of the users, the data relative to the fuel typology and consumption, the relative cost, the operation hours and the thermal power installed are available. The experimental data gathered, have supplied the data of the thermal power in the building and the corresponding fuel consumption each minute. From this data is possible to evaluate the energetic requirement and the mean thermal power supplied in the heating season. Currently the study area presents user majority fed by natural gas (relative to the residential buildings only two thermal plants are fed by liquid fuel with low sulphur tenor and one by gas oil; for the schools only natural gas is used). After a census, it results that the basin of user is served by thermal plants for a total power installed of 31.500 kW. Relative to the annual consumption an amount of about 2440.000 Nmc of natural gas is obtained, on the basis of the fuel price and the total expense data available for every residential building. Moreover, starting from the latter and the efficiency of each plant, the annual thermal requirement of about 19.870 MW h results. 3. Methodology for the simulation of the thermal request of a typical residential building 3.1. The surveys The objective of this work is to evaluate the hourly thermal load of a residential building, in every day of the heating season. It is possible, subsequently, to deduce the total thermal load for all the residential buildings, as described in the following section. In particular the analysis starts from the data of the thermal power supplied by the production unit. Moreover, the temperature data, acquired by a meteorological station located close the area interested by the district heating, are available for the period 01/01/2003 to 30/04/2004. Starting from these data it’s possible to assess the mean external temperature. The surveys of the thermal power generated and the fuel consumption for all the monitoring period, were gathered on a particular residential building fed by natural gas, chosen as the typical user in the period 10/11/2003 to 17/11/ 2003, with an acquisition frequency of 1 min for all the measurements. On the basis of the supplied data the generation efficiency equal to 82.8% was obtained. To investigate the correlation between thermal power and external temperature, a parameter called ‘‘Grado Giorno’’ (‘‘Deegre Day’’, DD) was introduced. It is defined, in reference to the italian DPR 26 agosto 1993, n. 412, as the difference between 20 8C and the mean daily external temperature according to expression (1): DD ¼ 20 Tme
(1)
Consequently the correlation is true:
2. Characterization of the user
Q ¼ DD C
The users to serve consist of only residential premises, in particular 46 buildings: there are three schools and 43
where Q is the energy needed to maintain an inner temperature of 20 8C in the building, and C is the global heat loss (kJ/8C)
(2)
1436
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
equal to the product of the heated volume and the specific dispersion coefficient per unit of both heated volume and DD. 3.2. Determination of constant C On the basis of the experimental data it’s possible to evaluate the thermal energy amounts Q, supplied to the user, referring to each day of the monitoring period. Moreover, the daily mean external temperatures from 6 a.m. to 24 p.m. (Tme6–24) is known and the corresponding values of DD; therefore C is estimated, according to Eq. (2), for all the days in the surveys. In particular for the residential building analyzed there is a value of the constant C of about 1421.347 kJ/8C, as the average of the constant C values calculated for the daily data available in the survey period. 3.3. Determination of the normalized thermal load profile for the typical residential building Starting from the measured data of the thermal power supplied, the attention was then focused on the trend of the thermal power, in order to determine the normalized thermal load profile for the typical residential building. This distribution w is defined according to expression (3), where ¯ is the mean thermal power supplied throughout the W ¯ h is the mean of the thermal power measurement period and W supplied in relation to the same hour in the different days of the monitoring period. In Table 1 are summarized the values of the mean thermal power for every hour relative to the monitoring period. The mean thermal power supplied during the measurement period is ¯ ¼ 210:9 kW. Knowing W ¯ h and W, ¯ the weights distribution w, W
Fig. 1. The normalized load thermal distribution ðwÞ.
illustrated in Fig. 1, was obtained according to Eq. (3) ¯h W w¼ ¯ W
(3)
Once that distribution w is modified and optimized as discussed below, it allows to determine the hourly trend of the thermal load for any residential building, knowing only the mean thermal power seasonally supplied and determined on the basis of the consumption and the thermal power nowadays installed. 3.4. Determination of the inner temperature Moreover, on the basis of the experimental data it’s possible to estimate the inner temperature trend in relation to the hourly thermal power supplied and the external temperature. It allows to verify the present effective operation of the system to
Table 1 Values of the hourly mean thermal power in the measurement period Hour
11/11/2003 (kW)
12/11/2003 (kW)
14/11/2003 (kW)
15/11/2003 (kW)
¯ h (kW) W
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 0 45 668.7 483.5 220.4 6.1 0 0 0 0 0 0 121.8 687.5 481.9 427.4 417.6 355.4 12.8 0
0 0 0 0 0 47 672.1 502.1 233.7 6.1 0 0 0 0 0 0 119.5 681.1 495.5 448.3 426.4 379.3 0 0
0 0 0 0 0 43 663.1 525.6 235.7 5.7 0 0 0 0 0 0 119.1 679 518.6 439.3 432 370.3 13.6 0
0 0 0 0 0 45 667.5 518.2 232 6.3 0 0 0 0 0 0 127.1 686.5 498.5 452.3 427.5 367.7 14.9 0
0 0 0 0 0 45 667.9 507.4 230.4 6.1 0 0 0 0 0 0 121.9 683.5 498.6 441.8 425.9 368.2 10.3 0
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1437
guarantee the inner comfort. To this aim another parameter has been defined Eq. (4), named Degree Hour DH: DH ¼
Qhour C
(4)
where Qhour is the energy supplied to the building each hour. Because of: DH ¼ Ti Tmeh
(5)
where Ti is the inner temperature and Tmeh is the external hourly temperature supplied by the meteorological station and mediated on the survey days. By applying Eq. (5), it is possible ¯h to obtain the Ti trend corresponding to the thermal power W actually supplied. As it can be seen in Fig. 2, the trend is not uniform and presents high peaks of temperature, much higher than the comfort temperature equal to 20 2 8C requested by the user. The latter corresponds to high peaks of thermal power, which must be reduced to guarantee the inner comfort for a longer time and to have a lower maximum power. The successive step is therefore the optimization of the normalized thermal load profile. 3.5. Optimization To underline the attainment of the regime conditions, corresponding to an inner temperature to 20 2 8C, the hourly distribution of the thermal power WREGx , necessary to maintain the regime condition in the structure 24 h/24 h, was calculated. To obtain that, first the DH distribution is calculated in reference to the mean external hourly temperature Tmeh and 20 8C inner temperature. From the comparison between this DH distribution, and the one evaluated in reference to the measured data (Eq. (4)), the regime conditions attainment results at 8.00 p.m. (Fig. 3) and is ¯ h of 438.2 kW. characterized by DH equal to 10.3 8C and a W With a simple proportion, it’s possible to obtain the value of the mean hourly thermal power necessary to have 24 h/day the user
¯h Fig. 2. Trend of the inner temperature corresponding to the thermal power W actually supplied.
Fig. 3. Trend of the Degree Hour DH distribution actually supplied compared to the one correspondent to the Ti = 20 8C.
in regime conditions, according to expression (6): WREGx ¼ 438:2
DDx 10:3
(6)
Analysing the WREGx with the values of thermal power really ¯ h along with the inner temperature supplied to the building W actually achieved, it’s possible to recognize the attainment of the regime conditions also in the morning as evidenced in Table 2. ¯ h and WREGx trends in corresponTherefore comparing W dence of the temperature transitory before regimen condition attainment both in the morning and in the evening, the effect to bring the building to the regime conditions was compared to the Table 2 Attainment of the regime conditions in the morning and in the evening Hour
Ti
¯ hour W
DH with the real Ti
DH with Ti = 20 8C
WREGx
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
8.1 7.9 7.8 7.8 7.6 8.6 23.2 18.7 14.1 9.9 10.7 11.1 11.6 11.6 11.2 10.6 12.9 25.9 21.4 20.0 19.7 18.4 9.9 8.7
0 0 0 0 0 45.5 669.5 456.8 219.5 4.6 0 0 0 0 0 0 109.6 682.1 497.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0 1.1 15.7 10.7 5.2 0.1 0 0 0 0 0 0 2.6 16.0 11.7 10.3 10.0 8.7 0.3 0
12.0 12.1 12.2 12.2 12.4 12.5 12.5 12.0 11.0 10.2 9.3 8.9 8.4 8.4 8.8 9.4 9.7 10.2 10.3 10.3 10.2 10.3 10.4 11.3
509.4 515.8 517.9 520.1 529.6 531.8 533.9 512.6 470.1 432.9 395.6 376.5 355.2 355.2 375.4 399.9 412.6 432.9 439.2 438.2 436.0 440.3 442.4 481.8
1438
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
one necessary only to maintain the same building in such conditions. The difference in the thermal power peaks is significant. To avoid this and obtain a more uniform daily trend of the inner temperature, the attention was focused on the thermal transitories considered as the periods necessary to bring the inner conditions near those of comfort. The solution proposed involves the ignition of the system an hour before supplying to the building the energy necessary to the thermal inertia in a 1 h longer interval. In particular the methodology proposed, detailed below, is based on the following assumptions:
where the denominator is the amount of the DH, characterizing the thermal transitory, calculated in relation to the mean external temperature in the hourly 4.00– 7.00 p.m. For the transitory in the morning (5.00–8.00 a.m.) a constant Cm (Eq. (10)) is obtained: Qm ¼ 1055:288 kJ= C (10) ð20 Tme3h Þ3 to have a new DH, relative to the time of ignition, it has been made as explained: the DH relative to the 2nd and 3rd hour of the transitory, have been opportunely fixed, in such a way to have one more uniform distribution of the thermal power values in those hours; the DH amount in the 4th hour is determined starting from the value of the energy, to supply during the last hour, divided per the constant C characterizing the transitory both in the morning and in the evening: Cm ¼
the transitories in the morning and in the evening are considered separately; the same DH values for the first hour of ignition of each transitory are considered; with the DH actually supplied to the building and known the constant C, the amounts of energy supplied to the user in the morning, 5.00–8.00 a.m. Eq. (7) and in the evening, 4.00– 7.00 p.m. Eq. (8) are calculated: X Qc ¼ C ðDHold Þ ¼ 43; 075:912 kJ (7) Qm
X ¼C ðDHold Þ ¼ 31; 519:853 kJ
¯ e ¼ Q Q13 ¼ 13; 835:640 kJ Q ¯ m ¼ Q Q13 ¼ 11; 396:795 kJ Q
(8)
Qm ¼ 1426:354 kJ= C ð20 Tme3h Þ3
(12)
where Q1–3 is the sum of the energy amounts supplied in the first 3 h of the thermal transitory.
subsequently it’s possible to calculate the constant C for the two thermal transitories. For the evening transitory Eq. (9), the Global Heat Loss is determined like: Ce ¼
(11)
Optimizing this trend, it’s possible to have new distributions of the DH, the thermal power supplied and the W values (Table 3). Fig. 4 shows the more uniform trend of the inner temperature, obtained according to expression (5), to better satisfy the request of the inner comfort.
(9)
Table 3 Revised Degree Hour DH, thermal power and load thermal distribution (w) Hour
DHnew
DHold
Ti new
Ti old
New thermal power (kW)
Old thermal power (kW)
Weights distribution
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 1.1 8 10 10.8 5.2 0.1 0 0 0 0 0 2.5 8 10 9.7 10.3 10 8.7 0.3 0
0 0 0 0 0 1.1 15.7 10.7 5.2 0.1 0 0 0 0 0 0 2.6 16 11.7 10.3 10 8.7 0.3 0
8 7.9 7.8 7.8 8.6 15.5 17.5 18.7 15.9 10 10.7 11.2 11.7 11.7 11.2 13.1 18.3 19.8 19.4 20 19.7 18.4 9.9 8.7
8 7.9 7.8 7.8 7.6 8.6 23.2 18.7 14.1 9.9 10.7 11.2 11.7 11.7 11.2 10.6 12.9 25.9 21.4 20 19.7 18.4 9.9 8.7
0 0 0 0 45.5 340.3 425.4 459.4 219.5 4.6 0 0 0 0 0 106.3 340.3 425.4 412.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0 45.5 669.5 456.8 219.5 4.6 0 0 0 0 0 0 109.6 682.1 497.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0.21 1.53 2.04 2.18 1.05 0.02 0 0 0 0 0 0.51 1.63 2.04 1.98 2.10 2.04 1.78 0.06 0
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1439
Fig. 4. New Ti and old Ti trends.
Fig. 6. Revised thermal load normalized distribution ðwÞ.
¯ new trend (Fig. 5) have been determined, both The values of W for the hours of the morning and evening transitory, in reference to the data of 7.00 a.m. and 8.00 p.m., respectively, according to expression (13):
with the real value resulting from the users’ census. Therefore, with the consumptions real data, the methodology has been verified in terms of energy. Starting from the data of the meteorological station, it’s possible to obtain the weights necessary to evaluate the hourly thermal load for all the heating season. In particular on the basis of the DD and the constant C ¯ d is obtained according values, the mean daily thermal power W to Eq. (14):
¯ hx ¯ h20 W W ¼ ; DHhx DHh20
¯ hx ¯ h7 W W ¼ DHhx DHh7
(13)
Both the peaks of Ti and the correspondent thermal power distribution are eliminated, with a consequent reduction of the thermal power requested. ¯ new it’s possible to obtain the revised w From the values W distribution (Fig. 6). This distribution of the weights is representative of the thermal load for a generic user, and it can be used as a reference for the same user typology. 3.6. Validation of the procedure on the test user To verify this practice the thermal request for every day in the period analyzed has been determined. The wnew , the correspondent DD and the surveys of the mean hourly temperature supplied by the meteorological station, were used to evaluate the annual energetic request and to compare this
¯ d ¼ C DD W 24 3600
(14)
where C is equal to 1421.347 kJ/8C and DD is the Degree Day value for each day of the heating season. Using this procedure, an energy requirement equal to 869.975 kJ and a maximum peak of thermal power equal to 961 kW are obtained for the type residential building in the period 01/01/2003 to 30/04/2004. On the basis of the energy requirement calculated and the primary energy consumed, an annual efficiency of generation equal to 79% has been obtained which, compared with the measured efficiency of 77%, confirms the validity of the calculation procedure. 4. Application of the methodology to all the residential buildings
Fig. 5. New and old thermal power trends.
Starting from the available data, for every residential building, the annual total energy consumed and the total hours of operation (15 h/day for 208 days) were deducted. Moreover, for every residential building, the mean thermal power has been determined as the ratio between the primary energy consumed and the total hours of ignition. This thermal power, multiplied for the generation efficiency, gives the ¯ a necessary for each residential annual mean power W building. To have the mean thermal daily power for all the residential ¯ d for the heating season of buildings, the daily mean power W the typical residential building has been considered. Moreover, the parameter K, defined as the ratio between the ¯ d value for each day and the W ¯ new thermal power of the typical W
1440
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
residential building (209.5 kW) is introduced Eq. (15): ¯d W K¼ ¯ W
(15)
¯ tot With K values for every day of the heating season and the W ¯ a of each thermal power (calculated as the sum of the W residential building) it has been possible to evaluate the daily ¯ gtot . Moreover, the mean power of all the residential buildings W thermal load of every day of operation has been achieved ¯ gtot value starting from the weights distribution and the W Eq. (16): ¯ gtot wnewhx ¯ gxtot ¼ W W
(16)
5. Validation of the procedure for all of the residential buildings ¯ tot value equal to about The data gathered result in a W 5.550 kW evaluated on the basis of the annual mean power, corresponding to the primary energy really consumed, and a generation efficiency supposed equal to 85% and 70% for natural gas and gas oil thermal plants, respectively. Moreover, the total real consumptions are equal to 16,982.890 kW h (corresponding to a total amount of energy supplied to the users of about 14,067.886 kW h), with a thermal power nowadays installed of 29.448 kW. According to the methodology detailed, the calculation model supplies the trend of the hourly thermal load for each day ¯ tot equal to of the heating season, resulting in a total W
5.245 MW for the period 1st January 2003 to 30th December 2003. This value differs from the corresponding real data of about 5.5%, confirming the validity of the proposed methodology. 6. Simulation of the thermal demand for the schools Also in this study case the authors begin from the determination of a model characterized in reference to a real case considered as the typical user. The procedure followed is the same detailed for the residential building case. The total real consumptions relative to the schools are equal to 931.256 kW h (corresponding a total amount of energy supplied to the user of about 776.267 kJ); with a thermal power nowadays installed of 1.316 kW, a mean generation efficiency equal to 85% and a mean thermal power supplied of 428.9 kW arise. Initially the power effectively supplied in some days of normal operation of the heating system has been measured; therefore with the data of the meteorological station the C parameter has been calculated, as described previously. From the analysis of the surveys, some peaks of thermal power in correspondence of the system ignition were observed, caused by the building thermal inertia related to the ignition transitory. As in the case of the residential building, the problem was resolved anticipating the ignition by an hour and following the procedure already detailed. Furthermore the procedure was extended first to all the days of the heating season and consequentially to all the schools of the basin user.
Table 4 ¯ h, the Degree Hour DH and the inner temperature Values of the hourly mean thermal power W Hour
Wh surveys
Wh with 1 h in advance of turn on
DH from the survey power
DH with 1 h in advance of turn on
Tmeo
Ti
w of the school
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 0 0 545.1 320.7 251.9 244.3 221.8 160.4 107.2 83.3 146.9 64.9 71.5 47.5 62.5 72.1 83.6 65.3 43.9 7.2
0 0 0 0 0 178.2 267.3 289.6 231.7 244.3 221.8 160.3 146.9 146.9 146.9 64.9 71.5 47.5 62.5 72.1 83.6 65.3 43.9 7.2
0 0 0 0 0 0 24.5 14.4 11.3 11.0 10.0 7.2 4.8 3.7 6.6 2.9 3.2 2.1 2.8 3.2 3.7 2.9 2.0 0.3
0 0 0 0 0 8 12 13 10.4 11.0 10.0 7.2 6.6 6.6 6.6 3.0 3.2 2.1 2.8 3.2 3.7 2.9 2.0 0.3
0 0 0 0 0 5.2 5.2 5.6 5.8 8.1 9.7 9.9 13.3 13.9 13 11.2 9.9 9.4 8.9 8.7 8.4 8.1 8 7.6
0 0 0 0 0 13.2 17.2 18.6 17.8 19.2 19.7 17.1 19.9 20.5 19.6 14.2 13.1 11.5 11.7 11.9 12.1 11.0 9.9 7.9
0 0 0 0 0 0.88 1.31 1.42 1.14 1.20 1.09 0.79 0.72 0.72 0.72 0.32 0.35 0.23 0.31 0.35 0.41 0.32 0.22 0.03
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1441
6.1. Analysis of the measured data and determination of the constant C The surveys of the school considered as the type user have been acquired in the period 23/11/2004 to 25/11/2004. ¯ supplied is equal to 136.9 kW in The mean thermal power W reference to the ignition time 5.00–12.00 a.m. The values of the ¯ h are shown in Table 4. The mean thermal hourly power W constant C is equal to 846.521 kJ/8C. 6.2. Determination of the normalized thermal load profile for the typical school As observed in the case of the residential buildings, the ¯ h is in relation with the DH distribution. hourly mean power W ¯ h distribution in order to reduce the Therefore, to modify the W power peaks is equivalent to modify the corresponding distribution of the DH. In the case of the schools, comfort conditions are requested daily in the time 7.00 a.m.–3.00 p.m. Moreover, during the ignition transitory (7.00–9.00 a.m.) a thermal energy Q equal to 42,478.561 kJ is supplied; the methodology adopted is based on anticipating the ignition by 1 h, leaving unvaried the amount of energy supplied in the thermal transitory. In the interval 6.00–8.00 a.m., the values of the DH were imposed to reduce thermal power peaks, guaranteeing at the same time indoor conditions close to those of the comfort ¯ h to the 9.00 a.m. has demanded. The value of the relative W been calculated imposing that Q remains unvaried in the thermal transitory. It has been therefore necessary to calculate the value of constant Cm relative to the thermal transitory Eq. (17) and the energy relative to the values imposed of DH between 6.00 and 8.00 a.m. Cm ¼
Q ð20 Tme3ore Þ 3
¼ 978; 768:7 kJ= C
(17)
In the hours after the thermal transitory, the values of the thermal powers initially measured were maintained unvaried; the obtained results are summarised in Table 4 and shown in Fig. 7.
Fig. 7. New and old thermal power trends.
Fig. 8. New Ti and the old Ti trends.
¯h Subsequently the trend of Ti corresponding to the revised W distribution was estimated. Also in this case significant are the improvements obtained (Fig. 8) in consideration of a target of both 20 2 8C inner temperature achievement in the comfort period, and the reduction of the peaks in the thermal power request. Finally the weights distribution has been determined in ¯ equal to 203.4 kJ relative to the system operation reference to W time (7.00 a.m.–3.00 p.m.). Therefore, the model developed can be considered as reference in order to characterize all of the schools. Fig. 9 shows the w distribution obtained both for the residential buildings and the schools. 6.3. Application of the methodology to all the schools Starting from the available data for every school, the annual total energy consumed and the total hours of operation are deducted. Moreover, for each school, the mean thermal power it has been determined as the ratio between the primary energy consumed and the total hours of ignition. This thermal power, multiplied by the generation efficiency, gives the annual mean ¯ a necessary for each school. To have the mean thermal power W daily thermal power for all the schools, the daily mean power
Fig. 9. Comparison between thermal load normalized distributions ðwÞ.
1442
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
¯ g has been considered for the heating season of the typical W school, according to Eq. (14). Moreover, once the parameter K Eq. (15) for every day of the ¯ tot have been heating season and the thermal power W determined, it’s possible to evaluate the daily mean power of ¯ gtot . the schools W According to the methodology previously detailed, the calculation model supplies during the entire heating season a total energy request of about 596.627 kJ and a maximum peak of thermal power equal to 653 kW.
Such a model allows to evaluate with precision the peak of the thermal request that must be satisfied, considering the contemporaneity of the loads and their different typology, and a more correct sizing of the generation plant. The results obtained illustrate that the maximum thermal power requested is smaller than the one nowadays installed. The benefits, both in environmental and energetic terms, are satisfactory and render favourable the realization of the plant.
7. Conclusions
[1] J. Agrell, P. Bogetoft, Economic and environmental efficiency of district heating plants, Energy Policy 33 (2005) 1351–1362. [2] A. Benonysson, B. Bohm, H.F. Ravn, Operational optimization in a district heating system, Energy Conversion and Management 36 (1995) 297–314. [3] G. Bloomquist, Reducing air emission through district heating and cooling, District Energy 82 (1997) 26–32. [4] S. Consonni, Environmental benefits of district heating with cogeneration, Termotecnica 6 (1995) 69–78. [5] A. Gebremedhin, B. Moshfegh, Modelling and optimization of district heating and industrial energy system, International Journal of Energy Research 5 (2004) 411–422. [6] F. Hansen, District heating based on natural gas, Euroheat and Power Fernwarme International 26 (1997) 9–12. [7] H.V. Larsen, H. Paisson, B. Bohm, H.F. Ravn, Aggregated dynamic simulation model of district heating networks, Energy Conversion and Management 43 (2001) 995–1019. [8] H. Zhao, Optimum operation of a CHP-based district heating system, Euroheat and Power Fernwarme International 24 (1995) 618–622.
The present work has been proposed to analyze the benefits in terms of energetic and environmental saving, through the realization of the district heating, combined to a cogeneration unit fed by natural gas. The objective was to create a model that allows to determine the hourly trend of the daily thermal load of all of the users, for every day of the heating season. To this aim this work proposes a new simulation model of the hourly and daily trend of the thermal load, knowing only few parameters for each user, like the power installed in the thermal plant, the seasonal operation hours and the timetable of the heating service distribution, as well as the external temperature trend.
References
Implementation of a cogenerative district heating: Optimization of a simulation model for the thermal power demand L. Barelli *, G. Bidini, E.M. Pinchi Dipartimento di Ingegneria Industriale, Universita` degli Studi di Perugia, Via G. Duranti, 67-06125 Perugia, Italy Received 12 November 2005; received in revised form 17 March 2006; accepted 20 March 2006
Abstract The district heating set up with a cogeneration system, concurs to attain energetic, economic and ambient benefits. It also provides to citizens a new service. The project strategy is based on the idea of supplying a portion of the necessary thermal power through a combustion alternative engine in cogeneration modality. It’s also interesting to modulate the load with auxiliary boilers fed by natural gas. This solution allows to save primary energy, create a centralization of the energy production, which contributes to the problem of polluting emissions, through the decentralization of the sources. The first step to assess the technical-economic feasibility of a district heating system, based on a cogeneration plant, is to underline and to characterize the energetic request of the basin of user. The objective of the present work is to develop a model that yields an esteem of the hourly thermal load for every days of the heating season of a complex user, represented by a single neighbourhood. To do this, the present work proposes a new method of simulation of the daily and hourly thermal load trend, known only the value of the power installed in the thermal plant for every user, the seasonal hours of the burner operation and the timetable of the heating service distribution, more than the external mean daily temperature trend. The results obtained using this model, have been verified with the data of seasonal consumptions, confirming the validity of the proposed methodology.The above allows to determine, with more precision, the thermal request peak to satisfy, taking in consideration the contemporaneity of the loads, also of different typology, and to carry out a better sizing of the generation plant. # 2006 Elsevier B.V. All rights reserved. Keywords: District heating; Simulation of the thermal demand; Cogeneration
1. Introduction At this moment the situation in terms of energy and energetic resource, results critical. The stock of the fossil fuel decreases gradually and the increase of the international energy request has produced a search of objectives, ratified in the Kyoto protocol, such as the rational use of energy, the improvement of the process’s efficiency, the passage towards fuel with a minor contained of carbon. But, the principal objective is to reduce the atmospheric emissions which cause the greenhouse effect. The present work proposes to develop a model of the thermal daily and hourly load of a complex user such as a neighbourhood during the heating season. This to value the thermal power peak necessary to correctly size the generation plant supplying heat to the district heating system [1–8]. The
* Corresponding author. Tel.: +39 075 5853740; fax: +39 075 5853736. E-mail address: [email protected] (L. Barelli). 0378-7788/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.enbuild.2006.03.023
model gives the seasonal amount of the thermal energy necessary to the buildings of the basin of user, to guarantee to the users, in function of the trend of the external mean seasonal temperature, the comfort requested. Knowing the total request, pointing out the better configuration of the operation, it’s possible to give a realistic evaluation of the fuel consumption necessary, the corresponding level of the emission in the environment, the electrical energy production and the energetic saving attainable. The strategy chosen is foreseen to guarantee, with a combustion alternative engine in cogeneration modality, a thermal power satisfying the thermal request characterizing the user for all of the heating season, supplying the remaining power, when necessary, through traditional boilers. In particular the study relates to the buildings, of an area comprising some neighbourhoods in Perugia, a city in central Italy, in order to arrange a district heating system combined with a cogeneration plant. This to increase the saving of primary energy obtainable, to realize the centralization of the
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
Nomenclature C DD DH Ti Tme w
global heat loss Degree Day Degree Hour internal temperature mean daily external temperature weights distribution
energy generation and the delocalization of the polluting production, controlling at the same time the emissions using suitable control systems. As a first step a census of the user of this study case has been conduced; the majority of the population is composed by residential buildings and some schools. The collected data for every thermal plant of the user, concern the combustible typology used, the seasonal consumptions, the power currently installed, the ignition hours of the burner and those of the heating service. For every user typology, a reference model deduced from experimental data has been conceived, in order to characterize the request of the typical user. This involves the determination of the weights distribution in 24 h defined as the ratio between the thermal mean hourly power distributed (valued as the mean of the thermal hourly powers supplied in the days of the surveys) and the thermal mean power distributed, calculated starting from the available data of all the surveys realized in the hourly ignition of the heating system. Having determined such weights distribution, it’s possible to assess the hourly trend of the inner temperature truly guaranteed with the current thermal plant. Then the weights distribution were varied through suitable modifications in the heating system operation, implying the reduction of the thermal power peaks requested in the thermal transitory at the ignition system. The subsequent analysis of the hourly trend of the inner temperature, allows to verify the satisfaction of the thermal comfort requested. To optimize this procedure and to obtain a more uniform daily trend of the inner temperature, the thermal transitory, considered as the period necessary to have the regime conditions in the building, was investigated. A modified thermal profile of the request, which characterizes the typical user, is obtained. This methodology was subsequently extended to the entire heating season, utilizing the weights distribution previously obtained and creating the correlation between the external temperature trend and the thermal request. Calculation procedure was finally extended to all the typologies of users, obtaining the daily and hourly trend of the total thermal load. On this basis the methodology was verified, through the true consumption data and the generation efficiency (valued starting from the hours available data), showing the efficiency of the model proposed.
1435
residential buildings. For all of the users, the data relative to the fuel typology and consumption, the relative cost, the operation hours and the thermal power installed are available. The experimental data gathered, have supplied the data of the thermal power in the building and the corresponding fuel consumption each minute. From this data is possible to evaluate the energetic requirement and the mean thermal power supplied in the heating season. Currently the study area presents user majority fed by natural gas (relative to the residential buildings only two thermal plants are fed by liquid fuel with low sulphur tenor and one by gas oil; for the schools only natural gas is used). After a census, it results that the basin of user is served by thermal plants for a total power installed of 31.500 kW. Relative to the annual consumption an amount of about 2440.000 Nmc of natural gas is obtained, on the basis of the fuel price and the total expense data available for every residential building. Moreover, starting from the latter and the efficiency of each plant, the annual thermal requirement of about 19.870 MW h results. 3. Methodology for the simulation of the thermal request of a typical residential building 3.1. The surveys The objective of this work is to evaluate the hourly thermal load of a residential building, in every day of the heating season. It is possible, subsequently, to deduce the total thermal load for all the residential buildings, as described in the following section. In particular the analysis starts from the data of the thermal power supplied by the production unit. Moreover, the temperature data, acquired by a meteorological station located close the area interested by the district heating, are available for the period 01/01/2003 to 30/04/2004. Starting from these data it’s possible to assess the mean external temperature. The surveys of the thermal power generated and the fuel consumption for all the monitoring period, were gathered on a particular residential building fed by natural gas, chosen as the typical user in the period 10/11/2003 to 17/11/ 2003, with an acquisition frequency of 1 min for all the measurements. On the basis of the supplied data the generation efficiency equal to 82.8% was obtained. To investigate the correlation between thermal power and external temperature, a parameter called ‘‘Grado Giorno’’ (‘‘Deegre Day’’, DD) was introduced. It is defined, in reference to the italian DPR 26 agosto 1993, n. 412, as the difference between 20 8C and the mean daily external temperature according to expression (1): DD ¼ 20 Tme
(1)
Consequently the correlation is true:
2. Characterization of the user
Q ¼ DD C
The users to serve consist of only residential premises, in particular 46 buildings: there are three schools and 43
where Q is the energy needed to maintain an inner temperature of 20 8C in the building, and C is the global heat loss (kJ/8C)
(2)
1436
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
equal to the product of the heated volume and the specific dispersion coefficient per unit of both heated volume and DD. 3.2. Determination of constant C On the basis of the experimental data it’s possible to evaluate the thermal energy amounts Q, supplied to the user, referring to each day of the monitoring period. Moreover, the daily mean external temperatures from 6 a.m. to 24 p.m. (Tme6–24) is known and the corresponding values of DD; therefore C is estimated, according to Eq. (2), for all the days in the surveys. In particular for the residential building analyzed there is a value of the constant C of about 1421.347 kJ/8C, as the average of the constant C values calculated for the daily data available in the survey period. 3.3. Determination of the normalized thermal load profile for the typical residential building Starting from the measured data of the thermal power supplied, the attention was then focused on the trend of the thermal power, in order to determine the normalized thermal load profile for the typical residential building. This distribution w is defined according to expression (3), where ¯ is the mean thermal power supplied throughout the W ¯ h is the mean of the thermal power measurement period and W supplied in relation to the same hour in the different days of the monitoring period. In Table 1 are summarized the values of the mean thermal power for every hour relative to the monitoring period. The mean thermal power supplied during the measurement period is ¯ ¼ 210:9 kW. Knowing W ¯ h and W, ¯ the weights distribution w, W
Fig. 1. The normalized load thermal distribution ðwÞ.
illustrated in Fig. 1, was obtained according to Eq. (3) ¯h W w¼ ¯ W
(3)
Once that distribution w is modified and optimized as discussed below, it allows to determine the hourly trend of the thermal load for any residential building, knowing only the mean thermal power seasonally supplied and determined on the basis of the consumption and the thermal power nowadays installed. 3.4. Determination of the inner temperature Moreover, on the basis of the experimental data it’s possible to estimate the inner temperature trend in relation to the hourly thermal power supplied and the external temperature. It allows to verify the present effective operation of the system to
Table 1 Values of the hourly mean thermal power in the measurement period Hour
11/11/2003 (kW)
12/11/2003 (kW)
14/11/2003 (kW)
15/11/2003 (kW)
¯ h (kW) W
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 0 45 668.7 483.5 220.4 6.1 0 0 0 0 0 0 121.8 687.5 481.9 427.4 417.6 355.4 12.8 0
0 0 0 0 0 47 672.1 502.1 233.7 6.1 0 0 0 0 0 0 119.5 681.1 495.5 448.3 426.4 379.3 0 0
0 0 0 0 0 43 663.1 525.6 235.7 5.7 0 0 0 0 0 0 119.1 679 518.6 439.3 432 370.3 13.6 0
0 0 0 0 0 45 667.5 518.2 232 6.3 0 0 0 0 0 0 127.1 686.5 498.5 452.3 427.5 367.7 14.9 0
0 0 0 0 0 45 667.9 507.4 230.4 6.1 0 0 0 0 0 0 121.9 683.5 498.6 441.8 425.9 368.2 10.3 0
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1437
guarantee the inner comfort. To this aim another parameter has been defined Eq. (4), named Degree Hour DH: DH ¼
Qhour C
(4)
where Qhour is the energy supplied to the building each hour. Because of: DH ¼ Ti Tmeh
(5)
where Ti is the inner temperature and Tmeh is the external hourly temperature supplied by the meteorological station and mediated on the survey days. By applying Eq. (5), it is possible ¯h to obtain the Ti trend corresponding to the thermal power W actually supplied. As it can be seen in Fig. 2, the trend is not uniform and presents high peaks of temperature, much higher than the comfort temperature equal to 20 2 8C requested by the user. The latter corresponds to high peaks of thermal power, which must be reduced to guarantee the inner comfort for a longer time and to have a lower maximum power. The successive step is therefore the optimization of the normalized thermal load profile. 3.5. Optimization To underline the attainment of the regime conditions, corresponding to an inner temperature to 20 2 8C, the hourly distribution of the thermal power WREGx , necessary to maintain the regime condition in the structure 24 h/24 h, was calculated. To obtain that, first the DH distribution is calculated in reference to the mean external hourly temperature Tmeh and 20 8C inner temperature. From the comparison between this DH distribution, and the one evaluated in reference to the measured data (Eq. (4)), the regime conditions attainment results at 8.00 p.m. (Fig. 3) and is ¯ h of 438.2 kW. characterized by DH equal to 10.3 8C and a W With a simple proportion, it’s possible to obtain the value of the mean hourly thermal power necessary to have 24 h/day the user
¯h Fig. 2. Trend of the inner temperature corresponding to the thermal power W actually supplied.
Fig. 3. Trend of the Degree Hour DH distribution actually supplied compared to the one correspondent to the Ti = 20 8C.
in regime conditions, according to expression (6): WREGx ¼ 438:2
DDx 10:3
(6)
Analysing the WREGx with the values of thermal power really ¯ h along with the inner temperature supplied to the building W actually achieved, it’s possible to recognize the attainment of the regime conditions also in the morning as evidenced in Table 2. ¯ h and WREGx trends in corresponTherefore comparing W dence of the temperature transitory before regimen condition attainment both in the morning and in the evening, the effect to bring the building to the regime conditions was compared to the Table 2 Attainment of the regime conditions in the morning and in the evening Hour
Ti
¯ hour W
DH with the real Ti
DH with Ti = 20 8C
WREGx
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
8.1 7.9 7.8 7.8 7.6 8.6 23.2 18.7 14.1 9.9 10.7 11.1 11.6 11.6 11.2 10.6 12.9 25.9 21.4 20.0 19.7 18.4 9.9 8.7
0 0 0 0 0 45.5 669.5 456.8 219.5 4.6 0 0 0 0 0 0 109.6 682.1 497.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0 1.1 15.7 10.7 5.2 0.1 0 0 0 0 0 0 2.6 16.0 11.7 10.3 10.0 8.7 0.3 0
12.0 12.1 12.2 12.2 12.4 12.5 12.5 12.0 11.0 10.2 9.3 8.9 8.4 8.4 8.8 9.4 9.7 10.2 10.3 10.3 10.2 10.3 10.4 11.3
509.4 515.8 517.9 520.1 529.6 531.8 533.9 512.6 470.1 432.9 395.6 376.5 355.2 355.2 375.4 399.9 412.6 432.9 439.2 438.2 436.0 440.3 442.4 481.8
1438
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
one necessary only to maintain the same building in such conditions. The difference in the thermal power peaks is significant. To avoid this and obtain a more uniform daily trend of the inner temperature, the attention was focused on the thermal transitories considered as the periods necessary to bring the inner conditions near those of comfort. The solution proposed involves the ignition of the system an hour before supplying to the building the energy necessary to the thermal inertia in a 1 h longer interval. In particular the methodology proposed, detailed below, is based on the following assumptions:
where the denominator is the amount of the DH, characterizing the thermal transitory, calculated in relation to the mean external temperature in the hourly 4.00– 7.00 p.m. For the transitory in the morning (5.00–8.00 a.m.) a constant Cm (Eq. (10)) is obtained: Qm ¼ 1055:288 kJ= C (10) ð20 Tme3h Þ3 to have a new DH, relative to the time of ignition, it has been made as explained: the DH relative to the 2nd and 3rd hour of the transitory, have been opportunely fixed, in such a way to have one more uniform distribution of the thermal power values in those hours; the DH amount in the 4th hour is determined starting from the value of the energy, to supply during the last hour, divided per the constant C characterizing the transitory both in the morning and in the evening: Cm ¼
the transitories in the morning and in the evening are considered separately; the same DH values for the first hour of ignition of each transitory are considered; with the DH actually supplied to the building and known the constant C, the amounts of energy supplied to the user in the morning, 5.00–8.00 a.m. Eq. (7) and in the evening, 4.00– 7.00 p.m. Eq. (8) are calculated: X Qc ¼ C ðDHold Þ ¼ 43; 075:912 kJ (7) Qm
X ¼C ðDHold Þ ¼ 31; 519:853 kJ
¯ e ¼ Q Q13 ¼ 13; 835:640 kJ Q ¯ m ¼ Q Q13 ¼ 11; 396:795 kJ Q
(8)
Qm ¼ 1426:354 kJ= C ð20 Tme3h Þ3
(12)
where Q1–3 is the sum of the energy amounts supplied in the first 3 h of the thermal transitory.
subsequently it’s possible to calculate the constant C for the two thermal transitories. For the evening transitory Eq. (9), the Global Heat Loss is determined like: Ce ¼
(11)
Optimizing this trend, it’s possible to have new distributions of the DH, the thermal power supplied and the W values (Table 3). Fig. 4 shows the more uniform trend of the inner temperature, obtained according to expression (5), to better satisfy the request of the inner comfort.
(9)
Table 3 Revised Degree Hour DH, thermal power and load thermal distribution (w) Hour
DHnew
DHold
Ti new
Ti old
New thermal power (kW)
Old thermal power (kW)
Weights distribution
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 1.1 8 10 10.8 5.2 0.1 0 0 0 0 0 2.5 8 10 9.7 10.3 10 8.7 0.3 0
0 0 0 0 0 1.1 15.7 10.7 5.2 0.1 0 0 0 0 0 0 2.6 16 11.7 10.3 10 8.7 0.3 0
8 7.9 7.8 7.8 8.6 15.5 17.5 18.7 15.9 10 10.7 11.2 11.7 11.7 11.2 13.1 18.3 19.8 19.4 20 19.7 18.4 9.9 8.7
8 7.9 7.8 7.8 7.6 8.6 23.2 18.7 14.1 9.9 10.7 11.2 11.7 11.7 11.2 10.6 12.9 25.9 21.4 20 19.7 18.4 9.9 8.7
0 0 0 0 45.5 340.3 425.4 459.4 219.5 4.6 0 0 0 0 0 106.3 340.3 425.4 412.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0 45.5 669.5 456.8 219.5 4.6 0 0 0 0 0 0 109.6 682.1 497.6 438.2 424.5 371.5 13.0 0.2
0 0 0 0 0.21 1.53 2.04 2.18 1.05 0.02 0 0 0 0 0 0.51 1.63 2.04 1.98 2.10 2.04 1.78 0.06 0
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1439
Fig. 4. New Ti and old Ti trends.
Fig. 6. Revised thermal load normalized distribution ðwÞ.
¯ new trend (Fig. 5) have been determined, both The values of W for the hours of the morning and evening transitory, in reference to the data of 7.00 a.m. and 8.00 p.m., respectively, according to expression (13):
with the real value resulting from the users’ census. Therefore, with the consumptions real data, the methodology has been verified in terms of energy. Starting from the data of the meteorological station, it’s possible to obtain the weights necessary to evaluate the hourly thermal load for all the heating season. In particular on the basis of the DD and the constant C ¯ d is obtained according values, the mean daily thermal power W to Eq. (14):
¯ hx ¯ h20 W W ¼ ; DHhx DHh20
¯ hx ¯ h7 W W ¼ DHhx DHh7
(13)
Both the peaks of Ti and the correspondent thermal power distribution are eliminated, with a consequent reduction of the thermal power requested. ¯ new it’s possible to obtain the revised w From the values W distribution (Fig. 6). This distribution of the weights is representative of the thermal load for a generic user, and it can be used as a reference for the same user typology. 3.6. Validation of the procedure on the test user To verify this practice the thermal request for every day in the period analyzed has been determined. The wnew , the correspondent DD and the surveys of the mean hourly temperature supplied by the meteorological station, were used to evaluate the annual energetic request and to compare this
¯ d ¼ C DD W 24 3600
(14)
where C is equal to 1421.347 kJ/8C and DD is the Degree Day value for each day of the heating season. Using this procedure, an energy requirement equal to 869.975 kJ and a maximum peak of thermal power equal to 961 kW are obtained for the type residential building in the period 01/01/2003 to 30/04/2004. On the basis of the energy requirement calculated and the primary energy consumed, an annual efficiency of generation equal to 79% has been obtained which, compared with the measured efficiency of 77%, confirms the validity of the calculation procedure. 4. Application of the methodology to all the residential buildings
Fig. 5. New and old thermal power trends.
Starting from the available data, for every residential building, the annual total energy consumed and the total hours of operation (15 h/day for 208 days) were deducted. Moreover, for every residential building, the mean thermal power has been determined as the ratio between the primary energy consumed and the total hours of ignition. This thermal power, multiplied for the generation efficiency, gives the ¯ a necessary for each residential annual mean power W building. To have the mean thermal daily power for all the residential ¯ d for the heating season of buildings, the daily mean power W the typical residential building has been considered. Moreover, the parameter K, defined as the ratio between the ¯ d value for each day and the W ¯ new thermal power of the typical W
1440
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
residential building (209.5 kW) is introduced Eq. (15): ¯d W K¼ ¯ W
(15)
¯ tot With K values for every day of the heating season and the W ¯ a of each thermal power (calculated as the sum of the W residential building) it has been possible to evaluate the daily ¯ gtot . Moreover, the mean power of all the residential buildings W thermal load of every day of operation has been achieved ¯ gtot value starting from the weights distribution and the W Eq. (16): ¯ gtot wnewhx ¯ gxtot ¼ W W
(16)
5. Validation of the procedure for all of the residential buildings ¯ tot value equal to about The data gathered result in a W 5.550 kW evaluated on the basis of the annual mean power, corresponding to the primary energy really consumed, and a generation efficiency supposed equal to 85% and 70% for natural gas and gas oil thermal plants, respectively. Moreover, the total real consumptions are equal to 16,982.890 kW h (corresponding to a total amount of energy supplied to the users of about 14,067.886 kW h), with a thermal power nowadays installed of 29.448 kW. According to the methodology detailed, the calculation model supplies the trend of the hourly thermal load for each day ¯ tot equal to of the heating season, resulting in a total W
5.245 MW for the period 1st January 2003 to 30th December 2003. This value differs from the corresponding real data of about 5.5%, confirming the validity of the proposed methodology. 6. Simulation of the thermal demand for the schools Also in this study case the authors begin from the determination of a model characterized in reference to a real case considered as the typical user. The procedure followed is the same detailed for the residential building case. The total real consumptions relative to the schools are equal to 931.256 kW h (corresponding a total amount of energy supplied to the user of about 776.267 kJ); with a thermal power nowadays installed of 1.316 kW, a mean generation efficiency equal to 85% and a mean thermal power supplied of 428.9 kW arise. Initially the power effectively supplied in some days of normal operation of the heating system has been measured; therefore with the data of the meteorological station the C parameter has been calculated, as described previously. From the analysis of the surveys, some peaks of thermal power in correspondence of the system ignition were observed, caused by the building thermal inertia related to the ignition transitory. As in the case of the residential building, the problem was resolved anticipating the ignition by an hour and following the procedure already detailed. Furthermore the procedure was extended first to all the days of the heating season and consequentially to all the schools of the basin user.
Table 4 ¯ h, the Degree Hour DH and the inner temperature Values of the hourly mean thermal power W Hour
Wh surveys
Wh with 1 h in advance of turn on
DH from the survey power
DH with 1 h in advance of turn on
Tmeo
Ti
w of the school
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
0 0 0 0 0 0 545.1 320.7 251.9 244.3 221.8 160.4 107.2 83.3 146.9 64.9 71.5 47.5 62.5 72.1 83.6 65.3 43.9 7.2
0 0 0 0 0 178.2 267.3 289.6 231.7 244.3 221.8 160.3 146.9 146.9 146.9 64.9 71.5 47.5 62.5 72.1 83.6 65.3 43.9 7.2
0 0 0 0 0 0 24.5 14.4 11.3 11.0 10.0 7.2 4.8 3.7 6.6 2.9 3.2 2.1 2.8 3.2 3.7 2.9 2.0 0.3
0 0 0 0 0 8 12 13 10.4 11.0 10.0 7.2 6.6 6.6 6.6 3.0 3.2 2.1 2.8 3.2 3.7 2.9 2.0 0.3
0 0 0 0 0 5.2 5.2 5.6 5.8 8.1 9.7 9.9 13.3 13.9 13 11.2 9.9 9.4 8.9 8.7 8.4 8.1 8 7.6
0 0 0 0 0 13.2 17.2 18.6 17.8 19.2 19.7 17.1 19.9 20.5 19.6 14.2 13.1 11.5 11.7 11.9 12.1 11.0 9.9 7.9
0 0 0 0 0 0.88 1.31 1.42 1.14 1.20 1.09 0.79 0.72 0.72 0.72 0.32 0.35 0.23 0.31 0.35 0.41 0.32 0.22 0.03
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
1441
6.1. Analysis of the measured data and determination of the constant C The surveys of the school considered as the type user have been acquired in the period 23/11/2004 to 25/11/2004. ¯ supplied is equal to 136.9 kW in The mean thermal power W reference to the ignition time 5.00–12.00 a.m. The values of the ¯ h are shown in Table 4. The mean thermal hourly power W constant C is equal to 846.521 kJ/8C. 6.2. Determination of the normalized thermal load profile for the typical school As observed in the case of the residential buildings, the ¯ h is in relation with the DH distribution. hourly mean power W ¯ h distribution in order to reduce the Therefore, to modify the W power peaks is equivalent to modify the corresponding distribution of the DH. In the case of the schools, comfort conditions are requested daily in the time 7.00 a.m.–3.00 p.m. Moreover, during the ignition transitory (7.00–9.00 a.m.) a thermal energy Q equal to 42,478.561 kJ is supplied; the methodology adopted is based on anticipating the ignition by 1 h, leaving unvaried the amount of energy supplied in the thermal transitory. In the interval 6.00–8.00 a.m., the values of the DH were imposed to reduce thermal power peaks, guaranteeing at the same time indoor conditions close to those of the comfort ¯ h to the 9.00 a.m. has demanded. The value of the relative W been calculated imposing that Q remains unvaried in the thermal transitory. It has been therefore necessary to calculate the value of constant Cm relative to the thermal transitory Eq. (17) and the energy relative to the values imposed of DH between 6.00 and 8.00 a.m. Cm ¼
Q ð20 Tme3ore Þ 3
¼ 978; 768:7 kJ= C
(17)
In the hours after the thermal transitory, the values of the thermal powers initially measured were maintained unvaried; the obtained results are summarised in Table 4 and shown in Fig. 7.
Fig. 7. New and old thermal power trends.
Fig. 8. New Ti and the old Ti trends.
¯h Subsequently the trend of Ti corresponding to the revised W distribution was estimated. Also in this case significant are the improvements obtained (Fig. 8) in consideration of a target of both 20 2 8C inner temperature achievement in the comfort period, and the reduction of the peaks in the thermal power request. Finally the weights distribution has been determined in ¯ equal to 203.4 kJ relative to the system operation reference to W time (7.00 a.m.–3.00 p.m.). Therefore, the model developed can be considered as reference in order to characterize all of the schools. Fig. 9 shows the w distribution obtained both for the residential buildings and the schools. 6.3. Application of the methodology to all the schools Starting from the available data for every school, the annual total energy consumed and the total hours of operation are deducted. Moreover, for each school, the mean thermal power it has been determined as the ratio between the primary energy consumed and the total hours of ignition. This thermal power, multiplied by the generation efficiency, gives the annual mean ¯ a necessary for each school. To have the mean thermal power W daily thermal power for all the schools, the daily mean power
Fig. 9. Comparison between thermal load normalized distributions ðwÞ.
1442
L. Barelli et al. / Energy and Buildings 38 (2006) 1434–1442
¯ g has been considered for the heating season of the typical W school, according to Eq. (14). Moreover, once the parameter K Eq. (15) for every day of the ¯ tot have been heating season and the thermal power W determined, it’s possible to evaluate the daily mean power of ¯ gtot . the schools W According to the methodology previously detailed, the calculation model supplies during the entire heating season a total energy request of about 596.627 kJ and a maximum peak of thermal power equal to 653 kW.
Such a model allows to evaluate with precision the peak of the thermal request that must be satisfied, considering the contemporaneity of the loads and their different typology, and a more correct sizing of the generation plant. The results obtained illustrate that the maximum thermal power requested is smaller than the one nowadays installed. The benefits, both in environmental and energetic terms, are satisfactory and render favourable the realization of the plant.
7. Conclusions
[1] J. Agrell, P. Bogetoft, Economic and environmental efficiency of district heating plants, Energy Policy 33 (2005) 1351–1362. [2] A. Benonysson, B. Bohm, H.F. Ravn, Operational optimization in a district heating system, Energy Conversion and Management 36 (1995) 297–314. [3] G. Bloomquist, Reducing air emission through district heating and cooling, District Energy 82 (1997) 26–32. [4] S. Consonni, Environmental benefits of district heating with cogeneration, Termotecnica 6 (1995) 69–78. [5] A. Gebremedhin, B. Moshfegh, Modelling and optimization of district heating and industrial energy system, International Journal of Energy Research 5 (2004) 411–422. [6] F. Hansen, District heating based on natural gas, Euroheat and Power Fernwarme International 26 (1997) 9–12. [7] H.V. Larsen, H. Paisson, B. Bohm, H.F. Ravn, Aggregated dynamic simulation model of district heating networks, Energy Conversion and Management 43 (2001) 995–1019. [8] H. Zhao, Optimum operation of a CHP-based district heating system, Euroheat and Power Fernwarme International 24 (1995) 618–622.
The present work has been proposed to analyze the benefits in terms of energetic and environmental saving, through the realization of the district heating, combined to a cogeneration unit fed by natural gas. The objective was to create a model that allows to determine the hourly trend of the daily thermal load of all of the users, for every day of the heating season. To this aim this work proposes a new simulation model of the hourly and daily trend of the thermal load, knowing only few parameters for each user, like the power installed in the thermal plant, the seasonal operation hours and the timetable of the heating service distribution, as well as the external temperature trend.
References