Annals of Nuclear Energy 110 (2017) 31–35

Contents lists available at ScienceDirect

Annals of Nuclear Energy

journal homepage: www.elsevier.com/locate/anucene

The calculation of the reactivity by the telegraph equation

Daniel A.P. Palma a,⇑, Aquilino Senra Martinez b, Fernando Carvalho da Silva b
a
Comissão Nacional de Energia Nuclear, Rio de Janeiro, Brazil
b
COPPE/Universidade Federal do Rio de Janeiro – Programa de Engenharia Nuclear. Rio de Janeiro, RJ, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Reactivity is one of the most important quantities of a nuclear reactor because it is directly related to
Received 13 January 2017 reactor core control. In a recent paper, Altahhan et al. (2016) derived a telegraph model of the point reac-
Received in revised form 6 June 2017 tor kinetics equations considering variable separation and mono-energetic neutron approximation. In
Accepted 9 June 2017
this paper the inverse point kinetics equation based on the telegraph model was obtained. The power his-
tory was calculated by the partial derivatives method as proposed by Diaz et al. In order to assess reac-
tivity accuracy considering the approximations involved in this model, simulations were performed for
different exponential power and relaxation time. The results for reactivity as obtained from the proposed
inverse point reactor equation show that it has significant differences in a comparison with conventional
formalism, being more pronounced as the relaxation time increases.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction
During the operation of a nuclear power plant it is necessary to
monitor neutron behaviour in the reactor core (Aboanber and
Mhlawy, 2009). The time-dependent behaviour of the neutrons in
response to any change in the material composition is important
for the safe operation of the reactor. Transient changes can take
place during reactor start-up (Zhang et al., 2008) or shutdown or
as a result of accidental disturbances in the operation of the reac-
tor. They may result in the departure of the reactor from the critical
condition. Therefore, it is very important to predict the time-
dependent behaviour of the neutron population as induced by
changes in the neutron multiplication (Chen et al., 2015), (Nahla
and Zayed, 2010) and (Li et al., 2009).
Reactivity is one of the most important quantities of a nuclear
reactor that represents the deviation from the condition of critical-
ity, and in the design of a nuclear power plant this information
may be used for a deeper investigation of the plant design bases
that are related to reactor operation. Control systems and opera-
tional procedures are used to limit the rate of reactivity variation.
Transient events as seen in a nuclear reactor can be predicted
only through the modification of the neutron flux and, as a result,
it is possible to make a sufficiently precise forecast on the conse-
quences of the transients. It is enough to relate the magnitude of
the neutron flux, which varies in time, to the neutron population
in the core of a nuclear reactor (Henry, 1975). Point kinetics equa-
tions relate these parameters and thus allow a study of the tran-
⇑ Corresponding author.
E-mail address: dpalmaster@gmail.com (D.A.P. Palma).
sient situations that may occur in a nuclear reactor, and their
obtaining takes place from a sequence of approximations, done
from the neutron transport equation. Their obtaining can be
accomplished directly from the neutron transport equation, for
the neutron diffusion equation, or through a heuristic procedure,
according to Stacey (2007) and Henry (1975).
The reactivity can be predicted through the inverse point reac-
tor kinetics model. This model results from the separation of the
spatial dependence by assuming a time-independent spatial flux
shape separated from a time-dependent amplitude function.
In practice, the use of point kinetics equations takes place in the
so-called inverse kinetics where the reactivity is obtained from the
nuclear power history (Duderstadt and Hamilton, 1976). There are
only a few problems for which it is possible to obtain an exact ana-
lytical solution for neutron density, given a specific reactivity.
Indeed, it is frequently more appropriate to invert the problem
by calculating the reactivity that will determine the past behaviour
for the neutron density expressed from a direct relation with the
nuclear power. This procedure, according to Henry (1975), is more
aligned with the nuclear reactor control methodology.
Many papers have been published with methods to determine
the reactivity using the inverse point kinetics equation. Recently,
Palma et al. (2016) derived such an equation from the neutron
transport equation considering the time variation of the neutron
density current in the P1-approximation. The kinetics parameters
were consistently defined for a multi-group energy structure. In
another paper Altahhan et al. (2016) derived a telegraph model
of the point reactor kinetics considering a variable separation
and mono-energetic neutrons approximation. The last is a very

http://dx.doi.org/10.1016/j.anucene.2017.06.017
0306-4549/Ó 2017 Elsevier Ltd. All rights reserved.
32 D.A.P. Palma et al. / Annals of Nuclear Energy 110 (2017) 31–35

crude approximation that leads to very simple kinetics parameters It is easy to see that when s0 ! 0 and with no external neutrons
definitions, with poor accuracy. source inside the reactor, Eq. (1) reproduces the conventional point
The goal of this paper consists of obtaining the inverse point kinetics equations well established in the literature (Duderstadt
kinetics equation based on the telegraph model as proposed by and Hamilton, 1976):
Altahhan et al. (2016) in order to evaluate the reactivity accuracy
  X
considering the approximations involved in this method. dnðtÞ qb m
¼ nðtÞ þ ki C i ðtÞ: ð6Þ
dt K i¼1
2. Telegraph point reactor kinetics equations
In the next section an expression for reactivity from the inverse
The set of equations based on the telegraph model of point reac- TPRK will be obtained.
tor kinetics (TPRK) as proposed by Altahhan et al. (2016) for a finite
reactor can be written thus:
3. Calculation of reactivity from inverse kinetics


sð1  qÞ  s0 ð1  bÞ
s0 n ðtÞ þ
ð2Þ
þ 1 nð1Þ ðtÞ In this section an expression for reactivity, based on the tele-
K
  X X graph inverse point kinetic equations, Eqs. (1) and (2), will be pre-
qb m m
ki C i ðtÞ þ s0 ki C i ðtÞ þ qðtÞ
ð1Þ
¼ nðtÞ þ sented. Before, we will obtain the inverse kinetic equation derived
K i¼1 i¼1 from the conventional point kinetics equations, Eqs. (2) and (6). In
þ s0 qð1Þ ðtÞ; ð1Þ all cases, a single effective delayed group was considered, that is:
X
m

ð1Þ b b¼ bi ; ð7Þ
C i ðtÞ ¼ i nðtÞ  ki C i ðtÞ; ð2Þ i¼1
K
" #1
respecting the following initial conditions:
1X m
bi
k¼ ; ð8Þ
nð0Þ ¼ n0 ; b i¼1 ki

nð1Þ ðtÞ ¼ 0;
t¼0 ð3Þ with m being the number of the delayed neutrons precursor group.
C i ð0Þ ¼ C i0 ¼ Kbki i n0 ;
3.1. Conventional inverse kinetics
and where

s0 ¼ 3D
v
From the set of conventional point kinetics equations, it is pos-
ð4Þ sible to write the following expression for the time variation of the
s ¼ 1þLs02 B2 ;
reactivity in relation to the amplitude function (Díaz et al., 2007):
are respectively the relaxation time for an infinite medium and a K ð1Þ kb
1 qcon ðtÞ ¼ b þ n ðtÞ  HðtÞ; ð9Þ
finite medium, and ð1 þ L2 B2 Þ is the non-leakage probability for nðtÞ nðtÞ
neutrons, with all other parameters defined as follow: v is the neu-
where HðtÞ is the power history of the rector as follows:
tron velocity, D is the neutron diffusion coefficient, B2 is the reactor
geometrical buckling, K is the neutron mean generation time, L2 is Z Z 1
t
0 0 hn0 i kt 0 0
the diffusion area, b is the total delayed neutron fraction of the fis- HðtÞ ¼ nðt 0 Þekðt tÞ dt ¼ e þ nðt0 Þekðt tÞ dt : ð10Þ
1 k 0
sion neutrons, ki is the delayed neutron decay constant for the ith
delayed neutrons precursor group, C i ðtÞ is the neutrons precursor The power history – Eq. (10) – will be calculated in this paper
concentration in the ith group, q is the reactivity and qðtÞ is the according to the method proposed by Díaz et al. (2007). It was
external source strength inside the reactor. demonstrated that, having integrated by parts, and in successive
In their paper, Altahhan et al. (2016) obtained Eqs. (1) and (2) in times, the integral in Eq. (10) it is possible to re-write it based on
a different way to Niederauer PhD thesis (1967), which includes
the ‘n’ order derivatives for amplitude function nðnÞ ðtÞ, as follows:
different variables definitions. The relaxation time was introduced
Z 1
to correct the drawback of the current, adjusted instantaneously to 0 0 knðtÞ  nð1Þ ðtÞ
nðt0 Þekðt tÞ dt ¼ nðtÞ
the gradient of the flux in Flick’s Law. Relaxation time is the param- 0 k2 nðtÞ  nð2Þ ðtÞ
eter specifying the rate of nuclear reactor returns to the state of
knð0Þ  nð1Þ ð0Þ
equilibrium. Greater values for relaxation time mean that the  nð0Þekt : ð11Þ
nuclear reactor returns to the state of equilibrium very slowly. k2 nð0Þ  nð2Þ ð0Þ
In the paper by Altahhan et al. (2016) other variables are used, In replacing Eqs. (10) and (11) one obtains the power history
as related to mono-energetic neutrons and homogeneous medium that appears in Eq. (9):
approximations. They are: " #
kb hn0 i kt knðtÞ  nð1Þ ðtÞ knð0Þ  nð1Þ ð0Þ
L2 ¼ RDa HðtÞ ¼  e þ 2 nðtÞ  2 nð0Þekt :
nðtÞ k k nðtÞ  nð2Þ ðtÞ k nð0Þ  nð2Þ ð0Þ
v Ra ¼ lð1þL1 B Þ 2 2 ð12Þ
keff 1
q¼ keff ð5Þ
3.2. Telegraph inverse kinetics equation
keff ¼ k1
ð1þL2 B2 Þ

K ¼ v R1a k1 ¼ keffl ¼ ð1  qÞl; The inverse telegraph point reactor kinetics equation will be
obtained considering a single effective delayed group and no exter-
where l is the neutron lifetime, k1 is the infinite reactor multiplica- nal source of neutrons. In this case, Eq. (1) can be simplified and
tion factor and Ra is the macroscopic absorption cross section. written by:
 D.A.P. Palma et al. / Annals of Nuclear Energy 110 (2017) 31–35 33

Table 1 Table 3
Nuclear parameters used in simulations. Reactivity as a result of time, calculated for nðtÞ ¼ expð1:5tÞ and parameters for case 1.

Group Case 1: Altahhan et al. Case 2: Palma et al. t (s) qcon ðtÞ (pcm) – Eq. (9) qtel ðtÞ (pcm) – Eq. (16) qtel ðtÞ  qcon ðtÞ
(2016) (2016) (pcm)
ki (s1) bi (%) ki (s1) bi (%) 0 93.71 93.93 0.23
2 758.20 758.22 0.02
1 0.0127 0.0285 0.0128 0.0161
4 786.48 786.50 0.01
2 0.0317 0.15975 0.0315 0.1117
6 787.69 787.70 0.01
3 0.115 0.141 0.1242 0.1000
8 787.74 787.75 0.01
4 0.311 0.30525 0.3281 0.2099
10 787.74 787.75 0.01
5 1.4 0.096 1.4085 0.0818
12 787.74 787.75 0.01
6 3.87 0.0195 3.8314 0.0207
14 787.74 787.75 0.01
K(s) 5  (10)4 1.35  (10)5
16 787.74 787.75 0.01
beff (%) – Eq. (7) 0.750 0.540
18 787.74 787.75 0.01
ki (s1) – Eq. (8) 0.078 0.086
20 787.74 787.75 0.01
D(cm) 0.143 10.0
B2 (cm2) 1.7  (10)7 1.7  (10)7
v(cm/s) 5  (10)5 5  (10)5
Ra ðcm1 Þ 0.170 0.0139
s0 ðsÞ  sðsÞ – Eq. (4) 1.95  (10)6 1.36  (10)4

Table 2
Reactivity as a result of time, calculated for nðtÞ ¼ expð0:12353tÞ and parameters for
case 1.

t (s) qcon ðtÞ (pcm) – Eq. (9) qtel ðtÞ (pcm) – Eq. (16) qtel ðtÞ  qcon ðtÞ
(pcm)
0 262.44 262.13 0.31
2 20.74 20.50 0.24
4 140.65 140.84 0.19
6 248.41 248.56 0.16
8 320.36 320.50 0.13
10 368.41 368.53 0.12 Fig. 1. Reactivity as a result of time, calculated for nðtÞ ¼ expð0:12353tÞ and
12 400.49 400.60 0.11 parameters for case 2.
14 421.91 422.01 0.10
16 436.21 436.31 0.10
18 445.76 445.86 0.09
20 452.14 452.23 0.09
22 456.40 456.49 0.09
24 459.24 459.33 0.09
26 461.14 461.23 0.09
28 462.41 462.50 0.09
30 463.26 463.35 0.09


sð1  qtel Þ  s0 ð1  bÞ
s0 nð2Þ ðtÞ þ þ 1 nð1Þ ðtÞ
K
 
qtel  b
¼ nðtÞ þ kCðtÞ þ s0 kC ð1Þ ðtÞ: ð13Þ
K
Substituting Eq. (2), and considering a single effective delayed Fig. 2. Reactivity as a result of time, calculated for nðtÞ ¼ expð1:5tÞ and parameters
for case 2.
group in Eq. (13), and rearranging the terms in order to isolate
the telegraph reactivity denoted by qtel , leads to:


qtel   K þ s  s0 ð1  bÞ ð1Þ where the constants are denoted by:
 snð1Þ ðtÞ þ n ¼ s0 nð2Þ ðtÞ  n ðtÞ
K K
    A1 ¼ s 0 K
s0 k  1 s0 k A2 ¼ K þ s  s0 ð1  bÞ
þb nðtÞ þ k 1  CðtÞ: ð14Þ
K K ð17Þ
A3 ¼ bð1  s0 kÞ
From Eq. (2), and considering a single effective delayed group, it kbðs0 kKÞ
A4 ¼ K
:
is possible to write the precursor concentration CðtÞ as follow:
Z The results calculated for reactivity as based on the inverse tele-
b t
0 0 b
CðtÞ ¼ nðt 0 Þekðt tÞ dt ¼ HðtÞ: ð15Þ graph point reactor kinetics equation, Eq. (16) will be reported in
K 1 K the next section.
being the power history HðtÞ as given by Eq. (12).
Replacing Eq. (15) in Eq. (14) and isolating reactivity qtel it is 4. Results
possible to write the following expression:

 1 In this section we will evaluate the reactivity behaviour in a
qtel ðtÞ ¼ A1 nð2Þ ðtÞ þ A2 nð1Þ ðtÞ þ A3 nðtÞ þ A4 HðtÞ snð1Þ ðtÞ þ n ;
nuclear reactor using the conventional and telegraph inverse point
ð16Þ kinetics. The nuclear parameters considered are those found in
34 D.A.P. Palma et al. / Annals of Nuclear Energy 110 (2017) 31–35

Table 4
Analysis of the terms of qtel ðtÞ for the case in which nðtÞ ¼ expð0:12353tÞ.

tðsÞ Case 1 (pcm) – Altahhan et al. (2016) Case 2 (pcm) – Palma et al. (2016)
105 ðT 1 þT 2 Þ 105 T 3 105 T 4 105 ðT 1 þT 2 Þ 105 T 3 105 T 4
snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn

0 0.8 750.0 1018.3 0.0 540.2 101.1

5 0.8 750.0 556.1 0.0 540.2 54.9
15 0.8 750.0 387.6 0.0 540.2 38.7
20 0.8 750.0 326.3 0.0 540.2 33.0
35 0.8 750.0 303.9 0.0 540.2 31.0
30 0.8 750.0 295.8 0.0 540.2 30.3

Table 5
Analysis of the terms of qtel ðtÞ for the case in which nðtÞ ¼ expð1:5tÞ .

tðsÞ Case 1 (pcm) – Altahhan et al. (2016) Case 2 (pcm) – Palma et al. (2016)
5 5 5
10 ðT 1 þT 2 Þ 10 T 3 10 T 4 105 ðT 1 þT 2 Þ 105 T 3 105 T 4
snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn snð1Þ ðtÞþn

0 112.5 750.0 731.1 3.2 540.1 71.2

5 112.5 750.0 37.5 3.2 540.1 4.0
15 112.5 750.0 37.2 3.2 540.1 3.9
20 112.5 750.0 37.2 3.2 540.1 3.9
35 112.5 750.0 37.2 3.2 540.1 3.9
30 112.5 750.0 37.2 3.2 540.1 3.9

Altahhan et al. (2016) and Palma et al. (2016) and can be seen in
Table 1.
Tables 2 and 3 shows the reactivity behaviour for an amplitude
function given by nðtÞ ¼ expð0:12353tÞ (Suescun et al., 2013) and
nðtÞ ¼ expð1:5tÞ (Palma et al., 2016) respectively considering case
1 with the kinetic parameters shown in Table 1.
Based on Tables 2 and 3 it is possible to see that the reactivity
calculated from the telegraph inverse kinetics and from the con-
ventional formulation overlap one another with the deviation from
the conventional formulation after it stabilises being of 0.09 pcm in
Fig. 1and 0.01 pcm in Fig. 2. As the transient is less pronounced, the
variation in reactivity due to the variation in the amplitude func-
tion quickly stabilises as well with the telegraph formulation sys-
tematically overestimating the results calculated by the
conventional formulation. In this case, the relaxation time is small
106.
Figs. 1 and 2 shows reactivity behaviour for an amplitude func-
tion as given by nðtÞ ¼ expð0:12353tÞ (Suescun et al., 2013) and
nðtÞ ¼ expð1:5tÞ (Palma et al., 2016) respectively, considering case
2 with the kinetic parameters shown in Table 1. In this case, the
relaxation time is longer (104) than in case 1, due to the fact that
the value for neutron diffusion coefficient is proportional.
Based on Figs. 1 and 2 it is possible to see that the reactivity cal-
culated from the telegraph inverse kinetics overestimates the con-
ventional one systematically. In Fig. 1 the difference
jqtel ðtÞ  qcon ðtÞj  198:1pcm while jqmod ðtÞ  qcon ðtÞj  25:2pcm
(Palma et al., 2016) after the stabilization that occurs rapidly as
time goes by indicates that something in the telegraph formalism
could be improved.
An analysis is made of Eq. (16) to identify which terms govern
the behaviour of the telegraph inverse kinetics equation presented
in this paper. For that, the following terms were defined:
T 1 ¼ A1 nð2Þ ðtÞ;
T 2 ¼ A2 nð1Þ ðtÞ;
ð18Þ
T 3 ¼ A3 nðtÞ;
T 4 ¼ A4 HðtÞ:
Tables 4 and 5 show an analysis of the time evolution for the
terms that form qtel for the two cases presented in this paper.
From the results shown in Tables 4 and 5, it is possible to see
that despite the fact that term T 3 rules reactivity behaviour, it is
not possible to disregard any terms in the Eq. (16) in order to sim-
plify the expression. In the cases presented in this paper, in which
nðtÞ ¼ ext , the reactivity calculated from the telegraph inverse
kinetics can be obtained by replacing nðtÞ in Eq. (16) and writing
it as:
A1 x2 þ A2 x þ A3 A4
qtel ðtÞ ¼ þ HðtÞext : ð19Þ
sx þ 1 sx þ 1
Finally, the reactivity calculated from the telegraph inverse
kinetics reaches an asymptotic value as follows:
A1 x 2 þ A2 x þ A3
limqtel ðtÞ ¼ ; ð20Þ
t!1 sx þ 1
which confirms the behaviour shown in Figs. 1 and 2 explicitly.
5. Conclusions
In this paper a telegraph inverse point kinetics equation was
obtained for an uniform reactor of arbitrary shape and using a
mono-energetic neutron approximation. Using an uniform reactor
allows separating the neutron flux, using a separation of variables
technique.
The mono-energetic neutron approximation is a questionable
assumption that could lead to very poor accuracy in the kinetics
parameters defined in Eq. (5) for some practical applications.
The results for reactivity as obtained from the inverse point
reactor equation, shown in Section 4 indicate that it has significant
differences from conventional formalism. This difference is more
pronounced when the relaxation time is increased.
It is not possible to disregard any term in Eq. (16). Even with
more terms, Fig. 1 shows that the difference of 198 pcm in relation
to the conventional term in the asymptotic state indicates that
something can be improved, as taking into account kinetic param-
eters consistently defined for a multi-group energy structure, as
proposed by Nunes et al. (2015).
The differences could be because the energy-dependent effects
were not taken into account. It should be noted that delayed neu-
trons appear with a somewhat lower energy than prompt
neutrons.
 D.A.P. Palma et al. / Annals of Nuclear Energy 110 (2017) 31–35
Acknowledgements
This work was supported by Research Support Foundation of
the State of Rio de Janeiro (FAPERJ), Brazilian Council for Scientific
and Technological Development (CNPq) and the Brazilian Institute
of Science and Technology for Innovative Reactors (INCTRI).
References
Aboanber, A., Mhlawy, A.M., 2009. Solution of two-point kinetics equations for
reflected reactors using analytical inversion method (AIM). Prog. Nucl. Energy
51, 155–162.
Altahhan, M.R., Nagy, M.S., Abou-Gabal, H.H., Aboanber, A.E., 2016. Formulation of a
point reactor kinetics model based on the neutron telegraph equation. Ann.
Nucl. Energy 91, 176–188.
Chen, W., Xiao, H., Li, H., Chen, L., 2015. Explicit appropriate basis function method
for numerical solution of stiff systems. Ann. Nucl. Energy 75, 353–357.
Díaz, D.S., Martinez, A.S., da Silva, F.C., 2007. Formulation for the calculation of
reactivity without a nuclear power history. J. Nucl. Sci. Technol. 44 (9), 1149–
1155.
35
Duderstadt, J.J., Hamilton, L.J., 1976. Nuclear Reactor Analysis. John Wiley & Sons,
New York.
Henry, Allan F., 1975. Nuclear Reactor Analysis. The MIT Press, Cambridge,
Massachusetts and London.
Li, H., Chen, W., Luo, L., Zhu, Q., 2009. A new integral method for solving the point
reactor neutron kinetics equations. Ann. Nucl. Energy 36, 427–432.
Nahla, A.A., Zayed, E.M.E., 2010. Solution of the nonlinear point nuclear reactor
kinetics equations. Prog. Nucl. Energy 52, 743–746.
Niederauer, G.F., 1967. Neutron Kinetics based on the Equation of Telegraphy (PhD.
thesis). Iowa State University of Science and Technology. Retrospective Theses
and Dissertations. Paper 3416.
Nunes, A.L., Martinez, A.S., Silva, F.C., Palma, D.A.P., 2015. A new formulation for
point kinetics equations considering the time variation of neutron currents.
World J. Nucl. Sci. Technol. 5, 57–71.
Palma, Daniel A.P., Martinez, A.S., Nunes, A.L., 2016. Effect of the time variation in
neutron current density in the calculation of the reactivity. Ann. Nucl. Energy
96, 204–211.
Stacey, Weston M., 2007. Nuclear Reactor Analysis. Wiley-VCH GmbH & CO KGaA,
Weinheim, Germany.
Suescun, D.D., Rodriguez, J.A.S., Figueroa, J.H.J., 2013. Reactivity calculation using
the Euler-MacLaurin formula. Ann. Nucl. Energy 53, 104–108.
Zhang, F., Chen, W., Gui, X., 2008. Analytic method study of point-reactor kinetic
equation during cold start-up. Ann. Nucl. Energy 35, 746–749.