ScienceDirect

Available online at www.sciencedirect.com

ScienceDirect
Energy Procedia 00 (2017) 000–000
www.elsevier.com/locate/procedia
Availableonline
Available onlineatatwww.sciencedirect.com
www.sciencedirect.com
Energy Procedia 00 (2017) 000–000

ScienceDirect
ScienceDirect
www.elsevier.com/locate/procedia

Energy
EnergyProcedia
Procedia120 (2017) 000–000
00 (2017) 317–324
INFUB - 11th European Conference on Industrial Furnaces and Boilers, INFUB-11
www.elsevier.com/locate/procedia

An optimal
INFUB algorithm
- 11th European to assess
Conference the compliance
on Industrial withINFUB-11
Furnaces and Boilers, the T2s
An optimalrequirement
algorithmoftoWaste-to-Energy
assess the compliance facilities
with the T 2s
The 15th International Symposium on District Heating
a,b and Cooling
requirement
Federico Viganò of Waste-to-Energy
*, Francesco Magli
a,b,c, facilities
Assessing theFederico
Department feasibility
of Energy
a
– Politecnico diof
Consorzio Viganò b
using
Milano,
a,b,c,
LEAP, Via N. Bixio*,27/C, the heat
Via R. Lambruschini
Francesco demand-outdoor
4/A, Milan
Magli
Piacenza 29121, Italy
a,b 20156, Italy

temperatureDepartment
function
MatER (Materialsfor
c
a long-term
and Energy district
from Refuse) Research heat LEAP
Center, c/o Consorzio demand forecast
of Energy – Politecnico di Milano, Via R. Lambruschini 4/A, Milan 20156, Italy
a
b
Consorzio LEAP, Via N. Bixio 27/C, Piacenza 29121, Italy
a,b,ccMatER (Materials
I. Andrić *, A. Pina , P. Ferrão , J. Fournier ., B. Lacarrière , O. Le Correc
a and Energy from
a Refuse) Research Center,
b c
c/o Consorzio LEAP
Abstract
a
IN+ Center for Innovation, Technology and Policy Research - Instituto Superior Técnico, Av. Rovisco Pais 1, 1049-001 Lisbon, Portugal
b
The gas resulting Veolia Recherche
from the incineration & Innovation,
of waste must be291 Avenue
raised, Dreyfous
after Daniel,
the last 78520of
injection Limay, France air, in a controlled and
combustion
Abstract c
Département Systèmes Énergétiques et Environnement - IMT Atlantique, 4 rue Alfred Kastler, 44300 Nantes, France
homogeneous fashion and even under the most unfavorable conditions, to a temperature of at least 850 °C for at least two seconds
(Art. 50.2 Directive 2010/75/EU). This norm and its variations (i.e. 1,100 °C for 2 s if the chlorine content of the incinerated waste
The gas resulting from the incineration of waste must be raised, after the last injection of combustion air, in a controlled and
exceeds 1% by mass), called “T2s requirement”, oblige all Waste-to-Energy (WtE) plant operators to monitor the post-combustion
homogeneous fashion and even under the most unfavorable conditions, to a temperature of at least 850 °C for at least two seconds
conditions
Abstract and to turn on auxiliary burners in the occurrence of noncompliance with such a requirement. In a WtE boiler, the
(Art. 50.2 Directive 2010/75/EU). This norm and its variations (i.e. 1,100 °C for 2 s if the chlorine content of the incinerated waste
determination of the mean temperature reached by combustion gas in the post-combustion zone, after an ideal residence time of
exceeds 1% by mass), called “T2s requirement”, oblige all Waste-to-Energy (WtE) plant operators to monitor the post-combustion
2District
s, is carried out networks
by an algorithm implemented in thein Distributed Control
as oneSystem the (DCS) of the plant. Currently, since many
conditionsheating
and to turn on auxiliary are commonlyburnersaddressed
in the occurrence the literature
of noncompliance of with most
such effective solutions
a requirement. In aforWtEdecreasing
boiler, thethe
different algorithms
greenhouse gas are
emissions used, it appears
from the buildingthat further
sector.investigation
These systems on this subject is required.
determination of the mean temperature reached by combustion gas inrequire high investments
the post-combustion which
zone, afterareanreturned through time
ideal residence the heat
of
This work
Dueconsiders, as a case study,conditions
an existingand WtE boiler and, by means of a calibrated long-furnace model of the post-
2sales.
s, is carried tooutthebychanged climate
an algorithm implemented in thebuilding renovation
Distributed Controlpolicies, heat demand
System (DCS) in the Currently,
of the plant. future could decrease,
since many
combustion
prolonging zone, investigates
the investment all
return the possible
period. operating conditions as well as their connections with the monitored variables. The
different algorithms are used, it appears that further investigation on this subject is required.
most relevant
The work
main scope influences
of this on the T 2s temperature are highlighted and some control algorithms are proposed.
This considers, as paper
a caseis study,
to assess
an the feasibility
existing WtE of usingand,
boiler the heat demand
by means of –aoutdoor temperature
calibrated long-furnacefunction
model forof
heat
thedemand
post-
The resultsThe
forecast. so far obtained
district of show that the
Alvalade, T2s isinaffected
located Lisbonboth by boilerwas
(Portugal), load and gas-side casefouling inThe
the same way is and for the same
combustion zone, investigates all the possible operating conditions as wellused as aconnections
as their study.with thedistrict
monitored consisted
variables.ofThe
665
extent. Therefore,
buildings since
thatinfluences
vary in gas-side
both fouling in period
construction the post-combustion zone is an uncontrollable variable, boiler load
high)is and
not usable in as
most relevant on the T2s temperature areand typology.and
highlighted Three
someweather
control scenarios
algorithms(low, medium,
are proposed. three district
input variable of a reliable
renovation algorithm. Moreover, the results highlight theTosignificant role
the that canobtained
be played in the algorithm for were
the
The results soscenarios
far obtainedwere developed
show that the(shallow, intermediate,
T2s is affected deep).
both by boiler estimate
load and gas-side error,
fouling in the same heatwaydemand
and forvalues
the same
estimation
compared of the T 2s by the oxygen content in secondary flue gas.
with results
extent. Therefore, since from a dynamic
gas-side foulingheat demand
in the model, previously
post-combustion zone is developed and validated
an uncontrollable by boiler
variable, the authors.
load is not usable in as
©The
2017 The Authors.
results showed Published
that when by Elsevier
only weather Ltd.
change is considered,
input variable of a reliable algorithm. Moreover, the results highlight the
the margin of error
significant could
role that canbebeacceptable for algorithm
played in the some applications
for the
Peer-review
(the error ofinunder responsibility
annual demand wasof lower
the organizing20%committee of INFUB-11.
estimation the T2s by the oxygen contentthan
in secondaryfor all
flueweather
gas. scenarios considered). However, after introducing renovation
©scenarios,
2017 Thethe error value
Authors. increased
Published up to 59.5%
by Elsevier Ltd. (depending on the weather and renovation scenarios combination considered).
©The
2017 The Authors.
value
Peer-review of
underslope Published
coefficientby
responsibility Elsevier
increased
of Ltd.average
on
the organizing withinofthe
committee range of 3.8% up to 8% per decade, that corresponds to the
INFUB-11.
Peer-review
decrease inunder responsibility
the number of heatingof the organizing
hours of 22-139hcommittee
duringoftheINFUB-11
heating season (depending on the combination of weather and
renovation scenarios considered). On the other hand, function intercept increased for 7.8-12.7% per decade (depending on the
coupled scenarios). The values suggested could be used to modify the function parameters for the scenarios considered, and
improve the accuracy of heat demand estimations.
* Corresponding author. Tel.: +39-02-2399-3886; fax: +39-02-2399-3913.
© E-mail
2017 The Authors.
address: Published by Elsevier Ltd.
[email protected]
Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and
* Corresponding author. Tel.: +39-02-2399-3886; fax: +39-02-2399-3913.
Cooling.
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
E-mail address: [email protected]
Peer-review under responsibility of the organizing committee of INFUB-11.
Keywords: Heat demand; Forecast; Climate change
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the organizing committee of INFUB-11.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd.

Peer-review under responsibility of the Scientific Committee of The 15th International Symposium on District Heating and Cooling.
1876-6102 © 2017 The Authors. Published by Elsevier Ltd.
Peer-review under responsibility of the organizing committee of INFUB-11
10.1016/j.egypro.2017.07.225
318 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324
2 Author name / Energy Procedia 00 (2017) 000–000

Keywords: Waste-to-Energy; T2s requirement; temperature after two seconds; suction pyrometer; boiler model; control algorithm.

1. Introduction

The gas resulting from the incineration of waste must be raised, after the last injection of combustion air, in a
controlled and homogeneous fashion and even under the most unfavorable conditions, to a temperature of at least
850 °C for at least two seconds [1]. This norm and its variations (i.e. 1,100 °C for 2 s if the chlorine content of the
incinerated waste exceeds 1% by mass), called “T2s requirement”, oblige all Waste-to-Energy (WtE) plant operators
to monitor the post-combustion conditions and to turn on auxiliary burners in the occurrence of noncompliance with
such a requirement. In a WtE boiler, the determination of the mean temperature reached by combustion gas in the post-
combustion zone, after an ideal residence time of 2 s, is carried out by an algorithm implemented in the Distributed
Control System (DCS) of the plant. Currently, no standard algorithm structure exists. A survey of the used methods
was carried out and showed that different variables are measured and used for the estimation of the T2s temperature,
as well as algorithm parameters vary from plant to plant and even among the identical incineration lines of the same
WtE plant. Table 1 summarizes the variables used by the algorithms adopted in five WtE plants.

Table 1. Variables used by the algorithms adopted in five WtE plants.

Plant 1 Plant 2 Plant 3 Plant 4 Plant 5
Measured temperatures X X X X X
Boiler load X X X X
Flue gas flowrate at stack X
Concentration of oxygen in flue gas X X

The T2s temperature is normally estimated by means of such a sort of algorithm, which is typically calibrated by
means of experimental procedures (like the one adopted in Germany [2]), based on the measurement of temperature
in the post-combustion zone with suction pyrometers. However, the measurements cannot be presumed to be
representative of any working condition of the boiler, since they are normally carried out in a limited set of conditions.
The only operating parameter whose influence is explicitly investigated during the measurements is usually the boiler
load (i.e. the steam production).
The T2s temperature primarily depends on the mean temperature of flue gas in the post-combustion zone and on the
flowrate of such gas. More schematically, it can be related to the flue gas flowrate and to its temperature at the entrance
of the post-combustion zone. However, both these data are not easily retrievable: the flowrate is usually measured
only at the stack of the plant, with some minutes of delay and after the unmeasured dilution operated by the ambient
air introduced along the gas path; the temperature can be measured with a reasonable accuracy only by means of
laboratory instruments, like suction pyrometers, and for limited time.
To warrant the compliance with the norm, during WtE plant operation, the T2s temperature must be estimated with
a sufficient safety margin based on easily available data, like the boiler load, the concentration of oxygen, measured
temperatures, etc. The preference of most operators is to limit as much as possible the number of variables involved,
by considering only those directly measured on the boiler, rather than those measured on other plant sections (like the
flue gas flowrate, normally measured only at the stack).
The aim of this work is to investigate the dependence of the T2s temperature of a WtE boiler from controllable (e.g.
boiler load, oxygen concentration) and exogenous (e.g. LHV of the burned waste, boiler fouling) variables, to define
possible algorithms that, based on few easily available operating data, can estimate with sufficient reliability the T2s
temperature. The sought result, considering also the effect of the exogenous factors, should always be a conservative
estimate of the actual T2s, featuring a sufficient safety margin with respect to the normative requirement.

2. Case study

The boiler of an existing WtE plant, on which some measurements have been recently carried out, has been
considered as a case study for this work. This allowed calibrating all the parameters of the model later described
 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324 319
Author name / Energy Procedia 00 (2017) 000–000 3

against real data. The calibration concerned the heat loss coefficient, geometrical factors of the combustion chamber,
fouling coefficients for the gas-side of the boiler.
The considered plant features three large incineration lines with water-cooled grate combustors, dual stage
combustion, pre-heated primary combustion air, semi-dry flue gas treatment system. It is fed only with pre-treated
municipal solid waste and was originally designed for RDF (Refuse Derived Fuel) processing.

3. Boiler model

The simulation of flue gas cooling along the post-combustion zone of a WtE boiler requires several models:

 a waste model, in terms of chemical composition and energy content; it allows determining the air requirement, the
flue gas production and the energy release during combustion;
 an overall lumped model of the boiler, in terms of overall mass and energy balances; this model basically sets the
connection between waste throughput and steam production;
 a model of the primary combustion zone, immediately above the grate; it is needed to determine the amount,
composition and temperature of the flue gas produced in such a zone, as well as the thermal power exchanged with
the combustion chamber walls;
 a model of the secondary combustion zone, which allows calculating the final composition of flue gas, as well as
their temperature after mixing and reacting with the secondary combustion air;
 a model of the post-combustion zone; it is needed to evaluate flue gas and walls mean temperatures along the gas
path.

The thermo-chemical properties of flue gas have been calculated by means of NASA polynomials [3], whereas its
emissivity has been calculated according to the Leckner’s method [4]. The corresponding gas absorptivity has been
calculated by means of the Hottel’s rule [5]. Finally, the effects of the typical high fly ash load in flue gas has been
considered adopting the approach proposed in the VDI Heat Atlas [6].

3.1. Waste model

The waste model has been adopted from Fellner et al. [7]. Their work showed that, for the purpose of mass and
energy balances, waste can be considered as composed of four components:

 ash, that is the dry inorganic part of the waste, which can be considered inert during combustion;
 moisture, that behaves as liquid water, absorbing its latent heat of evaporation during combustion;
 dry biogenic matter, i.e. the combustible fraction of biogenic origin – it features a limited LHV and as much limited
consumption of oxygen and production of flue gas per unit mass during combustion;
 dry fossil matter, that is the combustible fraction of fossil origin – it comprises mainly plastic materials, which
feature rather relevant LHV and as much relevant consumption of oxygen and production of flue gas per unit mass
during combustion.

Since each component features a certain LHV, as well as specific consumption of oxygen and production of flue
gas, practically any waste can be modelled by changing the shares of the four components. Table 2 summarizes the
specific properties adopted in this work for these four waste components.

Table 2. Properties of ash, moisture, dry biogenic matter and dry fossil matter adopted in this work.
Dry biogenic Dry fossil
Ash Moisture
matter matter
LHV, MJ/kg 0.0 -2.4425 18.5 36.0
Stoichiometric consumption of oxygen, Nm3/kg - - 0.9294 2.0485
Stoichiometric production of dry flue gas, Nm3/kg - - 4.3846 9.2304
Stoichiometric production of water vapour, Nm3/kg - - 0.7323 1.2173
320 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324
4 Author name / Energy Procedia 00 (2017) 000–000

3.2. Overall lumped model of the boiler

Fig. 1 depicts the scheme of the considered boiler enclosed in a control volume. The model entails the mass,
elemental and energy balances of the gas-side of the boiler, as well as the mass and energy balances of the water/steam-
side.
The mass and elemental balances of the gas-side connect the flowrates and compositions of the entering streams
(e.g. combustion air, waste) to the flowrates and compositions of the exiting streams (e.g. flue gas, bottom and fly
ashes). The energy balance of the same side gives the gross heat exchanged in the boiler.
Whilst the mass balance of the water/steam-side of the boiler is a simple relation between feedwater flowrate and
steam production (possibly accounting for the continuous blow down and other steam extractions), the energy balance
of this side gives the steam production, once steam parameters and feedwater conditions are set. The connection
between the energy balances of the two boiler sides is the gross heat exchanged, which determines the steam
production after the cutback due to boiler convection and radiation losses.

Control volume

Steam
drum
SH steam

Water b low down

Feedwater

To flue gas
treatment system
EVA SH ECO

EVA EVA

Refu se

Secondary air Boiler ash

Gr ate
combustor Water
Steam
Prim ary air Air
Flue g as

Bottom ash

Fig. 1. Scheme of the considered boiler, enclosed in a control volume.

3.3. Primary combustion model

Primary combustion takes place over the grate in conditions that can be, and usually are, sub-stoichiometric.
Therefore, significant amounts of gaseous species of incomplete combustion are produced. By adopting the simplest
two-step model of combustion [8], the only gaseous species of incomplete combustion considered is carbon monoxide
(CO). From the point of view of mass and elemental balances, the model is based on the application of stoichiometry.
Concerning the energy balance, instead, the Gurvich’s model for combustion chambers [9] has been adopted. This
approach condenses all the data describing geometry, flowrates and properties into few characteristic coefficients, to
allow calculating the ratio between the actual temperature of flue gas at the exit of the chamber and its corresponding
adiabatic temperature, hence the heat exchanged in the chamber. Other authors [10] have already used such a model
for WtE boilers. In this case, the only adaptation with respect to the original version of the model lies in the expression
 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324 321
Author name / Energy Procedia 00 (2017) 000–000 5

of the energy theoretically released during primary combustion. Since primary combustion can be sub-stoichiometric,
the heat release can be less than the LHV of the burned waste.

3.4. Secondary combustion model

Secondary combustion takes place at a certain height above the grate, where secondary combustion air is
introduced. To properly consider this phenomenon from the standpoint of mass and elemental balances, again
stoichiometry is applied, by assuming complete oxidation of all the gaseous species. The secondary combustion zone
corresponds to the first cell in the computation scheme use to model the post-combustion zone, as described in the
following. Therefore, its energy balance is solved according to such a scheme.

3.5. Post-combustion zone model

Since the crucial point in evaluating the behavior of the T2s is what happens in the post-combustion zone, its model
is the most sophisticated among those here discussed. It entails a 1-D scheme along the gas path, according to the
long-furnace approach proposed by Herapath and Peskett [11]. The peculiarity of this scheme is the explicitly
considered radiant heat exchange among consecutive gas elements. This is done by means of imaginary, transparent
and diffusive surfaces. The approach allows a fine discretization of the gas path, thanks to the consistency of the
method with regard to the length of the gas elements.
Special attention has been devoted to coupling this scheme with the Gurvich’s model for the combustion chamber.
The net radiation predicted by the latter on the imaginary surface dividing the combustion chamber from the post-
combustion zone has been matched with the difference between radiosity and irradiance of the same surface on the
post-combustion side.
By drawing the energy balance of each gas element in terms of absolute enthalpies, also the thermal effects of
chemical reactions have been considered. This is particularly true for the first cell of the post-combustion zone, where
secondary combustion takes place.

4. Investigated operating conditions

Since the operator of a WtE plant has many degrees of freedom in controlling the combustion process, and many
exogenous factors can influence it, a throughout survey of all the independent variables has been carried out. The
influence of ambient conditions and the degree of preheating of combustion air has been considered by means of four
possible sets of values, summarized in Table 3.

Table 3. Sets of values considered for ambient conditions and combustion air preheating.
Set 1 Set 2 Set 3 Set 4
Ambient temperature, °C 0.0 0.0 40.0 40.0
Ambient humidity, g/kgdry air 2.0 2.0 25.0 25.0
Primary air temperature, °C 5.0 120.0 45.0 120.0
Secondary air temperature, °C 5.0 5.0 45.0 45.0

For those input variables showing appreciable influence on the T2s, several values have been tested, whereas those
less influent have been set to values representative of normal operation. Table 4 lists all the considered variables with
the corresponding values so far tested: all their combinations have been used as simulation inputs in each of the four
sets of assumptions regarding ambient conditions and combustion air preheating. The total number of possible
combinations is almost 18,000, however, not all of them are feasible: as an example, the highest LHV value is not
compatible with the highest value of moisture content. The feasible simulations resulted circa 15,000.
322 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324
6 Author name / Energy Procedia 00 (2017) 000–000

Table 4. Operating variables that influence the T2s and their values used to carry out the simulations.
Input variable # of values Considered values
88.55 – 98.04 – 107.53 –
Boiler load (i.e. steam production), t/h 5
114.31 – 117.01
LHV of waste, MJ/kg 3 10 – 15 – 20
Moisture content of waste, % 3 15 – 30 – 35
Ash in the waste, kg/kgdry 1 0.18
Lambda factor of primary combustion 3 0.8 – 0.9 – 1
Grate cooling power, MW 1 1.7
4.5 – 4.7 – 5 – 5.3 – 5.5 – 5.7
Oxygen in the secondary flue gas, %Vol. 11
– 6 – 6.3 – 6.5 – 6.7 – 7
Fouling coefficient for the gas-side of the boiler, m2K/W 3 0.008 – 0.013 – 0.02
Flue gas temperature at boiler outlet, °C 1 175
Discharge temperature of bottom ash, °C 1 450
Mean discharge temperature of fly ash, °C 1 400
Unburnt carbon in bottom ash, kg/kgdry 1 0.01
Unburnt carbon in fly ash, kg/kgdry 1 0.03
Waste ash to bottom ash, % 1 90.0
Fly ash at the exit of the boiler / Total fly ash, % 1 66.7
Fouling factor in the Gurvich’s model 1 0.2
Total number of combinations 4,455 x 4

5. Results and discussion

The first outcome of our analysis concern the relation among simulation results and the temperature measured in
the post-combustion zone of the boilers by the plant instrumentation. Such a temperature is the main input variable
for the T2s estimation algorithm. In the case study, it is denoted as T24m, since it is measured at the plant elevation of
24 m. Given that flue gas temperature is higher than that of the internal surface of the boiler walls, the measured
temperature is somehow half way between these two temperatures.
The simulation of the 18 cases for which experimental data are available not only allowed calibrating the boiler
model, but also gave the possibility of comparing the measured T24m with the temperatures estimated by the model at
24 m elevation for flue gas and for the internal surface of boiler walls. The best fitting of the experimental data is
obtained when the measured temperature is represented by a weighted average of the two temperatures given by the
model. The two weights are 18% for the flue gas temperature and 82% for the temperature of the internal surface of
the boiler walls. The graph in Fig. 2 shows the comparison between measured (blue) and estimated (orange) T 24m for
all the cases: the estimated temperature resulted meanly 0.5 °C lower than the measured one, with a standard deviation
of circa 36 °C.
In the considered plant, flue gas temperature is measured at a certain distance from the last injection of combustion
air (i.e. the secondary air), corresponding to the plant elevation of 24 m. Such a position was probably chosen by the
boiler manufacturer to be as close as possible to the average position where the plane of 2 s ideal residence time is
found. However, even in the case the temperature is measured in a different location of the post-combustion zone, the
same procedure adopted for the T24m can be applied.
Once the measured temperature (i.e. the T24m in our case) can be estimated for each simulation, the resulting T2s
can be plotted against this variable. The graphs in Fig. 3 show the results, colored as functions of steam production in
plot (a), and of gas-side fouling coefficient in plot (b). The T2s is affected both by boiler load and gas-side fouling in
the same way and for the same extent. Consequently, since gas-side fouling in the post-combustion zone is an
uncontrollable variable, boiler load is not usable as input variable in a reliable algorithm.
The simplest algorithm to estimate the T2s as a function of T24m is the minimum envelope of the cloud of points in
the two graphs in Fig. 3. Therefore, a linear function is obtained:

T2 s C   0.4247  T24m C   578.9 (1)

 Federico Viganò et al. / Energy Procedia 120 (2017) 317–324 323
Author name / Energy Procedia 00 (2017) 000–000 7

Fig. 2. Comparison between flue gas mean temperatures at 24 m as measured (in blue) by the plant instrumentation and estimated (in orange) by
the model. The alphanumerical labels identify each measurement: the first digit refers to the number of the considered incineration line, the letter
P (Partial) and F (Full) specify the load condition, while, the last digit tells the chronological order of execution.

Fig. 3. T2s vs. T24m (both estimated) for all the simulations. Each point (representing one simulation) is colored depending on the value of a third
variable: steam production for plot (a), and gas-side fouling coefficient for plot (b).

By adopting the minimum envelope of the cloud of points, one is always in the safest side, since the estimated T2s
is always lower than the actual one. However, this can lead to switching on auxiliary burners even when the actual T2s
is much above the normative threshold, with waste of expensive auxiliary fuel. To improve the T2s estimation
accuracy, the effects of other monitored variables must be considered. A variable is deemed to be adapt in formulating
a control algorithm when it shows a clear effect on the minimum envelope of the cloud of points. This happens with
oxygen concentration in secondary flue gas (xO2), as it is shown in Fig. 4.
The minimum envelope of the cloud of points is always circa a straight line, with coefficients dependent on the
oxygen concentration xO2:

T2 s C    0.86  xO 2  0.4849  T24m C    2,175.5  xO 2  731.14 (2)

324
8 Federico
Author Viganò
name et al.
/ Energy / Energy00Procedia
Procedia 120 (2017) 317–324
(2017) 000–000

Fig. 4. T2s vs. T24m (both estimated) for all the simulations. Each point (representing one simulation) is colored depending on the corresponding
concentration of oxygen in secondary flue gas. The three solid lines represent the second proposed algorithm for xO2 = 5, 6, 7%Vol. on wet basis.

6. Conclusions

The results of this work show that the T2s is affected both by boiler load and gas-side fouling in the same way and
for the same extent. Consequently, since gas-side fouling in the post-combustion zone is an uncontrollable variable,
boiler load is not usable as input variable in a reliable algorithm.
Two algorithms are proposed for the case study considered: whilst the first one is a simple linear correlation
between the T2s and the T24m, with the risk of significant underestimation of the T2s, the second one improves
significantly the prediction by using also the oxygen concentration in secondary flue gas. Other extensive
investigations have been carried out by considering other monitored variables, however none of them influences the
correlation between the T2s and the T24m as clearly as the oxygen concentration in secondary flue gas does.

Acknowledgements

The authors thank dr. Gaboardi and dr. Giorgio for the contribution given through their Master Thesis work [12].

References

[1] European Parliament and Council. Directive 2010/75. O.J. L 334, 17.10.2010, p. 17-119.
[2] Bundesministerium für Umwelt Naturschutz und Reaktorsicherheit (BMU). Bundeseinheitliche Praxis bei der Überwachung der Emissionen –
RdSchr. D. BMU v. 13.06.2005 – Az.: IC I 2 – 45053/5. Published in: GMBI 2005 Nr.38, S.795.
[3] McBride BJ, Zehe MJ, Gordon S. NASA Glenn Coefficients for Calculating Thermodynamic Properties of Individual Species. NASA Tech
Rep 2002-211556.
[4] Modest MF. Radiative Heat Transfer. 3rd ed. Boston: Academic Press; 2013.
[5] Bergman TL, Incropera FP, DeWitt DP, Lavine AS. Fundamentals of Heat and Mass Transfer. 7th ed. Hoboken: Wiley; 2011.
[6] VDI e. V., editor. VDI Heat Atlas. 2nd ed. Berlin: Springer-Verlag Berlin Heidelberg; 2010.
[7] Fellner J, Cencic O, Rechberger H. A new method to determine the ratio of electricity production from fossil and biogenic sources in waste-to-
energy plants. Environ Sci Technol. 2007;41(7):2579–86.
[8] Westbrook CK, Dryer FL. Simplified Reaction Mechanisms for the Oxidation of Hydrocarbon Fuels in Flames. Combust Sci Technol.
1981;27(1–2):31–43.
[9] Kakaç S, editor. Boilers, Evaporators, and Condensers. Wiley-Interscience publication; 1991.
[10] Jegla Z, Bébar L, Pavlas M, Kropáč J, Stehlík P. Secondary combustion chamber with inbuilt heat transfer area - Thermal model for improved
waste-to-energy systems modelling. In: 19th International Congress of Chemical and Process Engineering, CHISA 2010 and 7 th European
Congress of Chemical Engineering, ECCE-7. Technicka 2, 616 69 Brno, Czech Republic; 2010.
[11] Herapath RG, Peskett S. Excess oxygen and temperature control in furnaces, a dynamic modelling study. J Inst Energy. 1987;60:171–84.
[12] Gaboardi M, Giorgio A. Il requisito T2s negli impianti di termovalorizzazione. Master Thesis (in Italian). Politecnico di Milano – School of
Industrial and Information Engineering, Milan, Italy; 2016.