Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 112 (2017) 555 562

Sustainable Solutions for Energy and Environment, EENVIRO 2016, 26-28 October 2016,
Bucharest, Romania

Computational model for a condensing boiler with finned tubes heat

exchanger
Nicolae Antonescua*, Paul-Dan Stnescua
a
UTCB Building Services Engineering Faculty Pache Protopopescu 66, Bucharest s2 , Romania

Abstract

European regulations state a minimum boiler efficiency for the main steady-state functioning conditions and for the overall
heating season conditions. In order to comply the demands of the new regulations, more and more boiler manufacturers are
switching the gas boilers production to condensing functioning. In the paper we present the numerical and computational
modeling of a boiler made out of packages of finned tubes and packages of smooth pipes. Validation of the computational results
is made by comparison with experimental results obtained in a certified laboratory for the modelled equipment (boiler). The
modelling was set on an aluminium finned tubes heat exchanger existent in the testing laboratory (part of a condensing boiler).
The geometry modeling considered the discretization of the heat exchanger in several zones, characterized by homogenous flue
gas flow speed, flow geometry and heat transfer characteristics (mainly feature-length). A good correlation between measured
and calculated values was obtained, the error ranging +/- 1% for the total and sensible heat and for thermal efficiency.
2017
2017TheTheAuthors.
Authors. Published
Published by Elsevier
by Elsevier Ltd. is an open access article under the CC BY-NC-ND license
Ltd. This
Peer-review under responsibility of the organizing committee of the international conference on Sustainable Solutions for Energy
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility
and Environment 2016. of the organizing committee of the international conference on Sustainable Solutions for Energy
and Environment 2016
Keywords: Heat and mass transfer, condensing boiler, computational model.

1. General context

The constant concern of the European countries for the energy efficiency lead to important legislation initiatives
also in the field of boiler efficiency [1]. So, European regulations state a minimum boiler efficiency for the main
steady-state functioning conditions and for the overall heating season conditions.

* Corresponding author. Tel.: +40742012987

E-mail address:nicuant@yahoo.com

1876-6102 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of the international conference on Sustainable Solutions for Energy and Environment 2016
doi:10.1016/j.egypro.2017.03.1116
556 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562

In order to comply the demands of the new regulations, more and more boiler manufacturers are switching the
gas boilers production to condensing functioning.
Condensing boilers can generally come either from an initial design as condensing boilers, either from normal
boilers fitted with modifications allowing the condensing functioning.

Nomenclature

rd and cv radiative and convective surface heat transfer coefficients , [W/(m2K)]

Tpw wall temperature (metallic pipe or insulation) , [K]
Tgm mean temperature of the flue gases , [K] ; T mg = (Tgi Tge) 0,5
Co10 8 Stefan Boltzmann constant for radiation
FL feature-length of the flow , [m]
Tt ; t t theoretical (adiabatic) burning temperature [K] ; [ oC ]
Tf ; t f furnace exit section flue gases temperature [K] ; [ oC ]
furnace temperature invariant; = Tf / Tt
Bo Boltzmann invariant for furnaces
M filling factor for the furnace (dimensionless)
af general absorption coefficient for the furnace
dirt coefficient
SR active radiative heat transfer surface of the furnace [ m2 ]
B fuel flow rate [ m3N / s ]
Vg technical (real) specific flue gases volume [ m3N / m3N ]
SCNV convective heat transfer surface [ m2 ]
ti ; te inlet and outlet temperature of the flue gases [ oC ]
ap radiation absorption coefficient for the heat transfer surface
ag radiation absorption coefficient for the flue gases
cpg specific heat capacity of the flue gases [J/( m3N K)] at a certain temperature [cpg = f (temp) ]
burning process efficiency ( 1)
Sh Sherwood invariant for mass transfer
DIF diffusivity coefficient
pH2O partial pressure of the water vapors (in the flue gases or at the wall temperature at saturation point)
In the first case, the design focuses from the beginning on special heat and mass transfer geometries characterized
by high flow speed for flue gases, small feature-length and combined gravitational and flue gas flow condensation
discharge. The burner is with premix of air and gas, with radiative
plane surface for stabilization and burning and it is placed over the
boilers body. The radiation component of the solid surface of the
burner is received by the finned pipes of the first row and is
considered as a solid-solid radiation
between two plane surfaces, one made of ceramics and one made of
steel (for the finned pipes, due to the diaphragm effect).
In the second case, the design targets mainly to use some existing
constructive parts from non-condensing boilers and fit them in a
condensing boiler. From that, it generally results the need for a
verification calculus for the chosen heat transfer surface when
functioning in condensing boiler conditions.
Usually, the manufacturers tend to operate modifications that
concerns mainly the general geometric characteristics of the heat
exchangers (pipes length, number of pipes per row, number of rows)
in order to obtain desires performance parameters without modifying the Figure 1 : boilers general design
 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562 557

intimate geometry of the heat exchanger (for example the fin geometry and the finned pipes positioning in the
package).
In the paper we present the numerical modeling of a boiler made out of packages of finned tubes and packages of
smooth pipes. The general design of the boiler can be seen in figure 1. The modeling work
targeted some specific results, as for example:
x Geometrical modeling of the boiler heat exchanger as existent in the reference solution (existent body in
production);
x Heat and mass transfer modelling in order to obtain a comprehensive set of functioning parameters (for a range of
thermal powers and condensing conditions);
x Validation of the computational results (and for the physical and computational model) by comparison with
experimental results obtained in a certified laboratory for the modelled equipment (boiler).

2. Computational results

The modelling was set on an aluminium finned tubes heat exchanger existent in the testing laboratory (part of a
condensing boiler). The geometry modelling considered the discretization of the heat exchanger in several zones, [2]
characterized by homogenous flue gas flow speed, flow geometry and heat transfer characteristics (mainly feature-
length). In figure 2 are presented the main zones considered for modelling discretization.

Upper part of the first row - R11

Lower part of the first row - R12

Second row of finned tubes - R2

Smooth pipes package - R3

Figure 2 the main zones considered for modelling discretization

The discretization process had as guidelines a number of considerations:

x the (discretization) element must be characterized by only one constructive solution;
x the element must be characterized by a single set of criterial equations;
x for an element the condensing state must be accepted as unitary, meaning with or without;
x the number of elements must be correlated with the desired complexity of the computational program;
x the boundary is set for the outer surface of the boiler body, including the burner;
x the inlet and outlet water temperature and the fuel flow rate are input parameters (along with all defining
physical, chemical and geometrical parameters) , the useful efficiency being the calculated parameter (along with
all related parameters);
558 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562

x the boiler is working in steady-state conditions and there are no heat losses through external surfaces (due to
constructive and functional particularities)

For the reference boiler the following zones were considered as heaving unitary thermal functioning by regard to
flue gas flow and heat (and mass if the case) transfer:
x the radiant surface of the burner, considered as being a premixed surface stabilization burner;
x the lateral surfaces of the furnace, generated by membrane-linked smooth pipes and the section of the finned
tubes heat exchanger considered as a cooled metallic plate (the radiation of the burner surface on the finned pipes
heat exchanger was considered as received by a smooth cooled metallic plate, due to the diaphragm effect of the
fins);
x first row of finned pipes;
x second row of finned pipes;
x the smooth pipes package.

Because from preliminary calculus resulted an important percentage of the total thermal load exchanged in the
first finned pipes row and because the surface extension geometry is variable for the upper and lower part of the
pipes (in the flue gas flow direction), it was decided to split the first finned pipes row in two separate zones (as
presented in figure 2).
So, as a result, the main heat exchange areas considered for the convective zone were:
x R11 upper part of the first row, were, because of the small height of the tube fins, the heat transfer is considered
governed by the flow over the exterior diameter of the pipe;
x R12 lower part of the first row were the flow is considered as being through narrow channels;
x R2 second row of finned pipes were the flow is also considered as being through narrow channels;
x R3 smooth pipes package were the flow geometry is characterized as transversal flow over exchequer pipes
arrangement.

The flow is considered to be through narrow channels if one of the dimensions of the flow cross section is
significantly bigger than the other one generating the possibility of considering the hydraulic equivalent diameter as
twice the channel small dimension. The flow direction dimension must be of the same order of magnitude as the
bigger dimension of the cross section (or bigger).
Modeling and calculation of sensible and latent (condensation) heat transfer for the reference solution (existing
boiler) was made aiming for the general characterization of boiler functioning by defining as main variables the
functional parameters such as temperatures, flow speeds and thermal loads [3]. Those parameters were determined
individually on each meshing surface, individualizing functioning parameters such as feature-length, physical
parameters of flue gases, convection and radiation coefficients of heat transfer, complex mass transfer coefficients,
surface extension efficiency etc.
The computational program constructed by the authors for the functional model of the boiler comprises the
following main parts:
x Input for constructive and functioning parameters (geometry, materials, combustible, burning process etc.);
x Thermal and physical parameters for the flue gases;
x Furnace heat exchange considering also the burner radiation [4] , [5] - (main equation used):

general characteristic equation for furnaces expressed via dimensionless criterions:

Bo 0, 6
T
M a f 0,6  Bo 0,6
 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562 559

verification equation for furnaces with no convective component:

Tt
tf  273
0,6
C 0 10 8 [ a f Tt3 S R
M 1
B V g c pg M

x Thermal verification for the convective heat exchangers (consecutive determination for the meshing surfaces, the
flue gases output parameters from the previous surface being input parameters for the following) using as main
equational base [4] , [5] more exactly the Timofeev criterial equations for the Nusselt coefficient :

convective equivalence of the flue gases radiation:

T 3,6
1  pw
a  1 Tgm
D RD 5,765 10 8
p 3
a g Tgm
2 T pw
1 
Tgm

verification equation for convective surfaces:

D CV D RD  S CNV

te 
t pw  ti  t pw e  B V g c pg M

x Determination of the total sensible heat flow transferred to the secondary agent (useful heat);
x Numerical determination of the condensing heat transfer for R2 (second finned pipes row);
o initial (input) flue gases temperature and composition;
o condensing heat transfer computation :
condensing condition for the wall temperature by regard to flue gases water
vapor saturation temperature;
calculus of the dimensionless criterions (invariants) specific for the mass
transfer process (Schmidt, Reynolds, Sherwood);
determination of the condensing coefficient :

Sh DIF 1 kg
J MASS 2,157 10  3 ( p H 2O  fluegases  p H 2O pipewall ) 2
FL t gm  273 m h

o convection coefficient determination for the secondary agent;

o calculus of the overall heat exchange coefficient considering:
sensible heat transfer from the flue gases to the wall;
condensing heat transfer from the flue gases to the wall;
surface extension geometry;
surface extension efficiency;
secondary agent heat transfer resistance
o computation loop closing errors with iteration reloop decision;
o final determination for the thermal and physical parameters;
560 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562

x Numerical determination of the condensing heat transfer for R3 (smooth pipes package);
o initial (input) flue gases temperature and composition;
o condensing heat transfer computation :
o convection coefficient determination for the secondary agent
o calculus of the overall heat exchange coefficient considering:
sensible heat transfer from the flue gases to the pipe;
condensing heat transfer from the flue gases to the pipe;
pipe wall thermal resistance
secondary agent heat transfer resistance
o computation loop closing errors with iteration reloop decision;
o final determination for the thermal and physical parameters;
x Final determination for the functioning parameters;
o sensible heat transferred to the secondary agent (useful heat);
o condensing useful heat flow (to the secondary agent);
o total useful heat flow;
o boilers thermal efficiency;
o flue gas temperature at exhaust coupling.

The program was run for the constructive solution measured in the Thermal Testing Laboratory belonging to the
Technical University of Construction, Building Services Engineering Faculty, for a range of thermal loads
(combustible debits) and for different secondary agent temperature values (for both condensing and non-condensing
conditions). The testing facility measuring error is less then 1% and the individual errors for the measuring elements
is 0,5%.
In table 1 are presented the experimental and computational results, along with the calculated errors, for the case
of condensing functioning of the boiler at nominal (maximal) thermal load.
The computation was performed for both nominal conditions as stated by the book and for nominal conditions
as realized on the testing facility (in order to achieve maximum comparison accuracy).
Table 1 : experimental and computational results at nominal (maximal) thermal load.
Base model Base model Base model Relative
boiler boiler boiler error
Values in brackets are input model results model results measured
values for the computational unit by the book testing values on the %
program setting facility setting testing facility
Inlet water temperature [C] ( 30 ) ( 31,2 ) 31,2
Outlet water temperature [C] ( 50 ) ( 51,6 ) 51,6
Combustible debit (gas network) [m3N / h] ( 2,571 ) ( 2,503 ) 2,503
Excess air at the burner [-] ( 1,25 ) ( 1,26 ) 1,26
Flue gas temperature [C] 55,5 55,5 47,9 0,46 *
Total useful thermal load [kW] 27,09 26,34 26,53 -0,72
Sensible useful thermal load [kW] 25,08 24,42 24,56 -0,57
Condensing useful thermal load [kW] 2,01 1,92 1,97 -2,54
Thermal efficiency (by respect to lower
[%] 105,9 105,8 106 -0,23
calorific power)
* error referred to the entire temperature span for the flue gases :
theoretical burning temperature to chimney temperature

From the data obtained and presented in table 1 some major primary conclusions can be drawn:

x there is a good correlation between measured and calculated values, the error ranging +/- 1% for the total and
sensible heat and for thermal efficiency;
 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562 561

x the 3% error range for the condensing thermal load is completely acceptable, moreover considering that in fact
the thermal load difference between model and experiment is about 50 W (0,05 kW) , a small value for a 24 kW
boiler, in the precision range of the measuring facility;
x although the flue gas temperature at the exhaust varies with approx 8 oC , value that might seem important for the
50 oC level of temperatures, but it must be considered the fact that the meaningful value is not the temperature
value per se but the temperature span for the flue gases; thats why we considered technically relevant to
calculate a relative error to the entire temperature span of the flue gases, meaning from the theoretical (burning)
temperature down to chimney (exhaust) temperature.

In order to verify the program (model) accuracy for different boiler working parameters, in table 2 are presented
the experimental and modelling results for another two operating regimes.

Table 2 : experimental and modelling results for another two boiler operating regimes
Relative Relative
Nominal 60/80 Nominal 40/60
error error
Values in brackets are input
values for the computational model measured model measured
unit % %
program results values results values

Inlet water temperature [C] ( 61,4 ) 61,4 ( 41,3 ) 41,3

Outlet water temperature [C] ( 81,1 ) 81,1 ( 61,1 ) 61,1
Combustible debit 3
[m N /h] ( 2,467 ) 2,467 ( 2,462 ) 2,462
(gas network)
Excess air at the burner [-] ( 1,25 ) 1,25 ( 1,24 ) 1,24
Flue gas temperature [C] 81,7 67,4 0,88 62,9 53,7 0,56
Total useful thermal load [kW] 23,77 24,37 -2,46 25,26 25,54 -1,10
Sensible useful
[kW] 23,77 24,37 -2,46 23,93 ---
thermal load
Condensing useful thermal
[kW] 0 0 1,33 ---
load
Thermal efficiency
(by respect to lower calorific [%] 96,8 97,3 -0,51 103,2 102,3 0,88
power)

Both operating regimes refer to nominal thermal power at burner level but are set to vary the condensing
conditions (as part of the total heat transfer), by eliminating the condensing conditions for the operating regime
Nominal 60/80 or by generating a limited condensing functioning for the operating regime Nominal 40/60.
By analyzing the results, the validation of the modelling program is confirmed for all the nominal power
operating conditions. The main comments linked to those secondary set of validation data are regarding mainly the
program accuracy evolution:

x the error in calculating the total thermal load increases along with the diminishing of the condensing heat
transfer, fact suggesting that probably (in the absence of a compensatory condensing heat flow) the sensible
heat transfer modelisation in the low temperatures flue gases area may be improved, possibly by considering the
turbulent effect of the upper heat transfer surfaces over the smooth pipes flue gas intake section;
x nevertheless, in all working conditions for nominal heat output, the modelling errors do not exceed 3%, value
considered as satisfactory when taken into consideration the complexity of the boiler heat exchange surface
geometry;

In order to complete the model and computational program validation, a partial thermal load operating regime
was also considered. The results are presented in table 3.
The conclusion by analyzing the data is that the model results fit very well the experimental results for all the
parameters : temperatures, thermal loads and efficiency.
562 Nicolae Antonescu and Paul-Dan Stnescu / Energy Procedia 112 (2017) 555 562

Table 3 : partial thermal load operating regime

Relative
Partial 40/60
error
Values in brackets are input values model
unit measured values %
for the computational program results
Inlet water temperature [C] ( 40,2 ) 40,2
Outlet water temperature [C] ( 58,9 ) 58,9
Combustible debit (gas network) [m3N /h] ( 1,478 ) 1,478
Excess air at the burner [-] ( 1,3 ) 1,3
Flue gas temperature [C] 46,1 47,3 -0,07
Total useful thermal load [kW] 15,37 15,45 -0,52
Sensible useful thermal load [kW] 14,48
Condensing useful thermal load [kW] 0,893
Thermal efficiency (by respect to Hi) [%] 102,7 103,0 -0,29

The variation of heat and mass transfer parameters due to the change of flow rates is correctly estimated by the
calculation model, fact that proves the validity of the criterial equations chosen for the model for the entire area of
operating regimes of the boiler.

3. Final conclusions

x The model is correctly structured both from the point of view of geometrical discretization and complex heat and
mass transfer;
x The physical phenomena of heat and mass transfer is correctly translated in criterial equations that prove their
validy on the entire span of operating regimes;
x Flue gases physical characteristics are correctly represented as numerical functions;
x Using the program for modelling close to the reference boiler geometries is possible;
x The computational analyze can offer information about the working parameters of each discretization surface,
generating the possibility of deeper functional analyze in order to improve the thermal and economical efficiency
of the boiler (from design stage).

References

[1] SR-EN harmonized standards for boiler construction, testing and labelling; performance parameters determination
[2] Mikael Naslund - On the design of residential condensing gas boilers - Doctoral thesis - January 1997 - Department of Heat and Power
Engineering - Lund Institute of Technology - ISSN 0282 1990 ISRN LUTMDN/TMVK 1011 SE - Printed in Sweden
[3] Danielle Makaire, Philippe Ngendakumana - Modelling the thermal efficiency of condensing boilers working in steady-state conditions -
Thermodynamics Laboratory Univeristy of LIEGE, Belgium
[4] Stanescu P.D., Antonescu N.N. Aparate Termice - Curs - Editura MATRIX Bucuresti 2013-978-973-755-878-7 432 pag.
[5] Antonescu N.N.- Instalaii de ardere i cazane cu eficien energetic ridicat i poluare redus Complemente de curs - Editura
MATRIX ROM Bucuresti 2011-ISBN 978-973- 755-699-8 269 pag.