Numerical Solution of A Parameter Estimation Probl
Numerical Solution of A Parameter Estimation Probl
Numerical Solution of A Parameter Estimation Probl
1. Introduction
For the disposal of low level (LLW) and intermediate level (ILW) radioactive waste in Ger-
many a final deep geological repository called Konrad is currently expected to go into
operation around the year 2030. In the process of disposal of nuclear wastes in Konrad one
of the major challenges are the compliance with waste acceptance criteria in order to assure
a safe and sustainable storage of the waste, and especially the proof of compliance. Radio-
logical waste acceptance criteria can be verified using passive measurements of the gamma
emitting radionuclide inventory of waste drums [1,2] and correlation of the remaining
radionuclides. For the waste acceptance criteria regarding the material (elemental) com-
position no metrological solution is applied as a standard so far requiring a recourse on
existing documentation or visual inspection by opening waste drums. To overcome this
situation, a measurement device called QUANTOM [3,4] has been developed that is capa-
ble of performing a non-destructive analysis of elemental masses of radioactive wastes in
200-l drums. The parameter estimation problem arising in the evaluation of the spatially
resolved measurement data from QUANTOM is investigated in this work.
neutron flux in the sample and the constant fraction of Avogadro constant NA and molar
mass Ml :
NA
PEγ ,l = · ml · Eγ ,l · σEγ ,l · . (1)
Ml
Hence, element masses can be obtained by inverting the proportionality. The computation
of neutron flux and the photopeak-efficiency will be discussed in detail in Section 2.
In more complex cases, where the dependence of the neutron flux on the elemental
composition can not be neglected, PGNAA evaluations are solved using an iterative for-
mulation. Hereby, the neutron flux for a given elemental composition is computed and,
using Equation (1), the updated elemental composition is determined. The element masses
in the (n + 1)-th iteration are thus computed as follows:
PEγ ,l Ml
(ml )n+1 = , (2)
NA · Eγ ,l · σEγ ,l · [(m1 )n , . . . , (mL )n )]
where L is the total number of elements. The iterative process terminates when the rel-
ative change of compositions between two iterations becomes neglectable. In [7,8] this
approach has been studied for the evaluation of one single PGNAA measurement using a
HPGe detector. Based on a relative formulation the authors were able to show the existence
of a solution under mild assumptions. Nevertheless, the uniqueness of a solution cannot
be guaranteed in this setting, although, in accompanying experimental studies ambigu-
ous solutions did not occur. The setup of our paper differs in several aspects from the one
considered in [7,8]. The sample volume in QUANTOM is larger than considered in stan-
dard literature before, hence spatial distribution of neutron flux and photopeak efficiencies
need to be carefully represented in the simulation. Measurements for the neutron flux
in QUANTOM present additional information regarding the reconstruction of elemental
compositions. Furthermore, the problem formulation of [8], where relative compositions
are determined, cannot be applied to the more complex measurement task of QUAN-
TOM. The model of evaluation in this work therefore aims at the determination of absolute
quantities.
Hence, we present an alternative to the standard iterative approach which is motivated
by the spatially resolved measurement and additional measurement equipment for the
neutron flux. We consider a mathematical formulation of the evaluation of QUANTOM
measurements as a parameter estimation problem. Based on this formulation, we make
use of gradient based methods to solve for the elemental composition. We only consider
the parameter estimation problem with respect to neutron transport model, the depen-
dence of photon transport to element composition is much weaker and can therefore be
neglected.
This problem is studied in a simplified setting which makes it possible to also address
the question of uniqueness of a solution. The material is assumed to be homogeneous, so
only one measurement position of the HPGe detector is considered. Both types of detec-
tors, for gamma spectra and total neutron flux, are considered in this setting. Elemental
compositions are represented by the vector ρ of all corresponding elemental densities ρl ,
which are parameters of the neutron transport equation and therefore a natural candidate
for the reconstruction. In this setting we are able to prove uniqueness of solutions to the
parameter estimation problem. To our knowledge, this is the first mathematically rigorous
4 A. JESSER ET AL.
result for PGNAA applications. Throughout this work we restrict the numerical setting to
two space dimensions. This is sufficient to study and present all major effects arising in the
problem setup. The problem setup is in parts comparable to Diffuse Optical Tomography
(DOT), as a similar governing equation is used as an approximate model for particle trans-
port. The connection between the two problems are discussed in the conclusive part of this
paper.
This paper is structured as follows. In Section 2 we introduce the SP1 -approximation of
the neutron transport equation and model functions for the detector response which, com-
bined, represent the forward model. Section 3 comprises a formulation of the parameter
estimation problem for QUANTOM in the form of a PDE-constrained optimization prob-
lem and a solution method via a formal Lagrange method. The key results of this work are
contained in Section 4, where we prove uniqueness of solutions to the parameter estima-
tion problem. In the following Section 5, we describe the computational implementation of
the parameter estimation problem, based on an already existing neutron transport solver.
Numerical results which support our findings are presented in Section 6, and we conclude
the paper in Section 7 by discussing both the generalizability of our results and remaining
open questions.
2. Forward model
As introduced in Section 1, two different kinds of detectors are used in QUANTOM to
detect the total neutron flux and the energy resolved emitted gammas, respectively. For
both detector types a model for the counting rates is required. Figure 2 shows the abstract
measurement setup. Neutrons are constantly emitted by a neutron source. As a result, a
steady state neutron flux field is formed in the graphite and the sample, while the graphite
moderator provides a thermal neutron spectrum both in the graphite and the sample. It
is important to note here, that the neutron flux field depends on the elemental composi-
tion of the sample, since the neutrons directly interact with its atoms. Both the sample and
graphite are activated by the free neutrons, resulting in element-specific gamma emissions.
The detectors are placed in proximity to the sample in order to collect the related measure-
ment signal. Activation of the graphite can be neglected since scattering is dominant in the
Figure 2. Abstract sketch of the measurement setup, including the neutron source, sample and both
types of detectors in a surrounding graphite moderator.
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 5
moderator. Furthermore, detectors can be collimated to the sample to further decrease this
effect.
In order to evaluate the counting rates of the detectors, it is necessary to introduce mod-
els for the gamma and photon transport first. The transport of gammas and neutrons can
be described by the linear Boltzmann equation [9]. For both types of transport we make
use of approximate models. For the neutron transport equation, the SP1 -equation, also
known as neutron diffusion equation, is used as an approximation. The applicability of the
SP1 -equation for this setup was shown in a previous publication [6]. Note that in this work
we apply the one-group SP1 -equation, since we can assume a thermal neutron flux. The
equation then reads
1
−∇ · ∇φ(x) + a (x)φ(x) = Q(x). (3)
3t (x)
with the Marshak boundary conditions
1 1 ∂
φ(x) + φ(x) = 0 (4)
2 3t (x) ∂n
where φ(x) is the neutron flux and Q(x) the source term while t (x) and a (x) are the
total and absorption cross sections of the neutrons. It is important to note that the total and
absorption cross sections within the sample can be expressed using the elemental densities
ρl as well as the total and absorption cross sections of the elements, σtl and σal , while these
cross sections are independent of ρ anywhere else, leading to the representation:
⎧
⎨ σtl ρl x ∈ Vsam
t (x) = l
⎩
Mt (x) else
⎧
⎨ σal ρl x ∈ Vsam
a (x) = l (5)
⎩
Ma (x) else
Here Mt (x) and Ma (x) are piecewise constant functions that represent the cross sections
in the actual facility. The dependence of the cross sections on ρ in the sample leads to a
non-linearity of this parameter estimation problem.
According to Equation (1), photon transport is relevant for the evaluation of a single
peak as it determines the so-called photo-peak efficiency Eγ ,l . This quantity represents
the fraction of gammas which are detected at full energy. Consequently, scattered photons
do not contribute to Eγ ,l , allowing relatively simple models to calculate photo-peak effi-
ciencies, rather than employing full photon transport simulations. The main effects which
have to be considered are absorption, the geometry of the measurement setup and detector
physics. Photopeak efficiency is a parameter dependent on gamma energy. Therefore, eval-
uating a measurement requires either point-wise values for all gamma energies considered
or determining a (gamma-)energy-dependent calibration curve for a given measurement
facility or setup, accounting for all these effects. This calibration curve provides photopeak
efficiencies for any given gamma-energy and can be obtained experimentally, by measur-
ing at given energies using calibration emitters and then interpolating between measured
6 A. JESSER ET AL.
values. Alternatively, calibration curves can be determined using either analytical mod-
els or Monte Carlo simulations. As prompt gamma energies range between 50 keV up to
12 MeV, experimental calibration is of limited use, making modeling or simulation-based
calibrations preferable. Optimal performance is typically achieved using Monte Carlo sim-
ulations with MCNP [10] to account for the intricate physics within a HPGe detector.
Further details on the calibration of gamma detectors can be found in [11].
In principle, the photopeak efficiency also depends on the composition of the sample,
which changes in every iteration of the evaluation (Equation (2)) (as the current result of
the measurement, not the physical composition!). However, the tabulated values in [12]
suggest that the mass attenuation coefficients for most of the materials relevant for the
application are similar. Therefore, the sample’s impact on the photo-peak efficiency is pri-
marily governed by its average density, a known parameter that remains constant during
the evaluation. Thus, we assume photo-peak efficiencies to be constant throughout the
iterative evaluation.
In this work, we employ a simplistic model for the photopeak efficiencies in order to
simplify notations and calculations, but all results remain valid when using experimentally
obtained calibration curves, more sophisticated analytical models, or calibration curves
based on extensive Monte Carlo simulations. The photopeak efficiency is given by
Eγ ,l (x) = geom,Eγ ,l (x) · att,Eγ ,l (x) · det,Eγ ,l , (6)
where att,Eγ ,l (x) is the photon attenuation, geom,Eγ ,l (x) is the geometric efficiency and
det,Eγ ,l is the detector efficiency. It should be noted that the photopeak efficiency is depen-
dent of the location x where the photon is emitted. Since there exists no closed analytical
expression for the full photopeak efficiency, we use a model that captures its main features.
The geometric efficiency in this setup is given by
1 rdet (x)
geom,Eγ ,l (x) = arctan , (7)
π ddet (x)
where rdet (x) is the radius of the detector and ddet (x) the distance to the center of the
detector. The photon attenuation is given by
att,Eγ ,l (x) = e−μ·d(x) , (8)
where µ is the attenuation coefficient and d the traversed distance. Hereby the attenuation
coefficient is mainly determined by the average density of the material, which can be easily
determined in practice. The detector efficiency det,Eγ ,l can be interpreted as the probability
for a photon entering the detector to contribute to the measured signal and has to be deter-
mined by photon transport simulations. For a given energy and detector, it is a constant.
For simplicity it is set to one in the following.
Using both the neutron flux φ(x) and the photopeak efficiency Eγ ,l , we can define the
counting rates of the detectors. The neutron detectors measure neutron counting rates
given by
Rn [φ] = σd · φ(x) dx, (9)
Vdet,n
where σd is the so-called detector cross section. The integral is carried out over the volume
of the nth detector Vdet,n .
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 7
In contrast to the neutron detectors, the gamma detectors do resolve the energy of the
measured photons. The resulting gamma spectrum can then be analyzed for element-
specific photopeaks. Assuming a homogeneous sample, the counting rate for photons
forming a photopeak at energy Eγ specific for element l is given by
PEγ ,l [ρ, φ] = Eγ ,l ρl σEγ ,l φ(x) dx, (10)
Vsam
where σEγ ,l is the so-called partial (n,γ )-cross-section and ρl is the density of the element
l within the sample. Equation (10) is a generalization of Equation (1), accounting for the
additional spatial variation of the parameters Eγ ,l and φ. The integral in Equation (10) is
carried out over the volume where the gamma photons are produced, the sample volume
Vsam .
L
wγ
N
wd
Jdet [ρ, φ] = (P [ρ, φ] − PEγ ,l )2 +
2 Eγ ,l 2
(Rn [φ] − Rn )2 . (11)
l=1 PEγ ,l n=1 Rn
Here, wγ and wd are weights that determine the contribution of the different detectors to
the objective function, while L is the number of elements within the sample and N the
number of neutron detectors. The weight of the neutron detectors is set to wd = 12 , while
the weight of the gamma detectors is set to wγ = 34 to ensure an equal weight of both terms
despite the higher number of neutron detectors. In addition, we include a term punishing
deviations from the known total mass M:
wm
Jm [ρ] = 2
(c1 ρ1 + c2 ρ2 − M)2 (12)
M
Here c1 and c2 are material-specific constants and wm = 12 . Since the total mass M can
be measured with high accuracy, we assume in the following that the error on the total
8 A. JESSER ET AL.
mass is negligible. This term effectively acts as a regularization. Its effect is investigated in
Section 6.2.
The complete objective function is then given by
The parameter estimation problem can then be formulated as minimizing the objective
function with the SP1 -equation as a constraint, so we want to solve the problem
while D is the domain with corresponding boundary ∂D. Outside the sample, t and a are
independent of ρ. Thus, the differential operator D depends on ρ on a part of the domain.
Since the operator D(x) is self-adjoint, i.e. D(x) = D (x), we can identify φ with the
adjoint flux. The boundary terms in Equation (17) need to vanish. This is only the case
if the relation
1 1 ∂
φ (x) + φ (x) = 0 (18)
2 3t (x) ∂n
is fulfilled on the boundary. The derivatives of the Lagrangian can be used within an itera-
tive scheme to reconstruct the elemental densities ρ. Differentiation of Equation (16) with
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 9
respect to the adjoint flux φ yields the weak form of the forward SP1 -equation, while the
differentiation with respect to the flux φ yields the weak form of the adjoint SP1 -equation
D (x)φ (x) = Q (x) (19)
with the adjoint source
δJ
Q = [ρ, φ]
δφ
L
2wγ
= (P [ρ, φ] − PEγ ,l )
2 Eγ ,l
Eγ ,l (x)ρl σEγ ,l dx
l=1 PEγ ,l Vsam
N
2wd
+ 2
(Rn [φ] − Rn ) σd dx. (20)
n=1 Rn Vdet,n
and the boundary condition defined in Equation (18). The derivative of Equation (16) with
respect to the density of the lth element is given by
δL σt,l
[ρ, φ, φ ] = ∇φ · ∇φ dx − σa,l φφ dx
δρl j σt,j ρj Vsam Vsam
2wγ
+ 2
σ ρ
Eγ ,l Eγ ,l l φ dx − PEγ ,l Eγ ,l σEγ ,l φ dx
PEγ ,l Vsam Vsam
2wm cl
+ 2
(c1 ρ1 + c2 ρ2 − M). (21)
M
Using the results for φ and φ , the gradient of the Lagrangian with respect to the different
elemental density ρl is computed and can be used in a gradient-based optimization scheme.
4. Uniqueness
In this chapter, we investigate the uniqueness of the solution of the unperturbed parame-
ter estimation problem, i.e. without noise in the measurement data. First, we introduce a
change of variables to simplify the discussion. We then investigate the objective function
to motivate our proof for the uniqueness result that we finally present.
of ρ. The parameter η in Equation (23) is only varied within Vsam and effectively reduces
there to
a
η = 2 = 3a t (24)
γ
due to the homogeneity of the sample. For the investigation of uniqueness, the term pun-
ishing deviations from the total mass M is not needed, so we set wm in Equation (12) to zero.
Using the definition of ψ in Equation (22), we then obtain a transformed cost functional
L
wγ
N
wd
J[ρ, ψ] = (P [ρ, ψ] − PEγ ,l )2 +
2 Eγ ,l 2
(Rn [ψ] − Rn )2 . (25)
l=1 PEγ ,l n=1 Rn
−(σa1 σt2 + σa2 σt1 )ρ1 + (σa1 σt2 + σa2 σt1 )2 ρ12 − 4(σa1 σt1 ρ12 − η/3)σa2 σt2
ρ2 =
2σa2 σt2
(28)
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 11
Figure 3. Subfigures (a)–(c) depict the different terms of the objective function after decomposition,
subfigure (d) depicts the full objective function. The analytically computed minimum of the objective
function of the neutron detectors is green. A logarithmic scale is used for all plots. (a) Objective func-
tion including the first Photopeak. (b) Objective function including the second Photopeak. (c) Objective
function including all neutron detectors and (d) Full objective Function.
If the value of η belonging to the minimum is used, Equation (28) can be used to compute
the expected position of the minimal values. It is depicted in Figure 3(c) as the green curve.
It can be seen that the green curve lies within the valley of the numerical minimal values.
From Figure 3(a)–(c) it is apparent that neither the measurements of the different
gamma lines nor the neutron flux measurements alone have a unique minimum. In combi-
nation however, we obtain an objective function with a minimum contained within a small
region, well suitable for a reconstruction, as depicted in Figure 3(d).
Theorem 4.1: Let J[ρ, ψ] be the objective function as defined in Equation (25), ψ being a
solution of Equation (23), whereas ρ, t , a > 0. Then J has a unique minimum.
12 A. JESSER ET AL.
The proof of Theorem 4.1 relies on the following two lemmas, which are consequences
of the weak and strong maximum principle, respectively.
Lemma 4.2: Let D be a bounded domain with smooth boundary ∂D. Let u be a weakly
differentiable function satisfying a mild regularity assumption (cf. Appendix A for details)
and
−u(x) + η(x)u(x) = Q(x) in D (29)
with bounded, non-negative η and Q that are strictly positive on a subset of D that is not a
null set. Then u is strictly positive in D.
Lemma 4.3: Let D be a bounded domain with smooth boundary ∂D. Then the problem
The proofs of both lemmas are carried out in Appendices A and B. We can now present
the proof of the main result.
Proof: In the following, the different terms in the objective function (Equation (25)) are
investigated separately. We define JP,min as the set of values ρ that minimizes the first term
in Equation (25) (corresponding to the gamma detectors) and JR,min as the set of values ρ
the minimizes the second term in Equation (25) (corresponding to the neutron detectors).
Since both terms are positive, the intersection of JP,min and JR,min is then the set of values
that minimizes the full objective function. The proof works as follows: showing that there
exists a unique solution ψmin that minimizes JR,min using monotonicity arguments, we find
explicit expressions for the elemental densities ρl which are again unique.
Since the ψ is determined by the parameter η, the neutron counting rates defined in
Equation (26) can be seen as implicitly dependent of the parameter η. In the following
we show that ψ and thus the counting rates are strictly monotone in η. We assume in the
following that we have two parameters with η̃ > η, and corresponding ψ̃ and ψ, such that
Equation (23) is satisfied:
Q
−∇ · ∇ψ + ηψ =
γ
Q
−∇ · ∇ ψ̃ + η̃ψ̃ = . (31)
γ
By subtracting Equation (31) and defining
λ = ψ − ψ̃ and δ = η − η̃ (32)
we obtain
−∇ · ∇λ + ηψ − η̃ψ̃ = 0, (33)
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 13
which is equivalent to
−∇ · ∇λ + η̃λ = −δψ. (34)
By applying Lemma 4.2, we obtain the strict positivity of ψ and since δ < 0, the strict
positivity of λ follows directly with the same lemma. Since λ is strictly positive, we have
ψ̃ < ψ for η̃ > η. The counting rate Rn [ψ] defined in Equation (26) is a positive linear
functional of ψ, thus we also obtain monotonicity of Rn [ψ]. This ensures that there is just
one ηmin minimizing Jn .
With Lemma 4.3 it follows directly that there is only one ψmin that minimizes the second
term in Equation (25), so that JR,min = {ρ|η(ρ) = ηmin }. Now we can compute the inter-
section of the sets of minima of both terms in Equation (25). Inserting ψmin and using the
homogeneity of the sample and the linearity of the integral, Equation (27) yields
ρl
PEγ ,l = IEγ ,l = ρl 3 σt,l ρl IEγ ,l (35)
γ l
where
IEγ ,l = Eγ ,l (x)σEγ ,l ψmin (x) dx. (36)
Vsam
From this system of L equations, where L is the number of elements, we can, using some
algebra, derive linear relations for between the different elemental densities, which are valid
for all pairs l , l with 1 ≤ l , l ≤ L:
By inserting them in Equation (35), we obtain for each element l the relation
3/2
PEγ ,l IEγ ,l
PEγ ,l = ρl 3 σt,l IEγ ,l . (38)
l
PEγ ,l IEγ ,l
Since all expressions apart from ρl are constant, it is apparent that there is exactly one value
for each ρl , 1 ≤ l ≤ L, that satisfies the equation. As a consequence, Jmin contains just one
point.
It should be noted that the argument would still be valid if a different model had
been chosen for the gamma counting rates defined in Equation (6). The generalization
to heterogeneous media is discussed in the concluding Section 7.
5. Implementation
For the neutron transport computations, a neutron transport code named SPARC [6] is
used. We apply the finite element method for the spatial discretization of Equation (3), to
handle the geometric complexity of the underlying model. Furthermore, the finite element
method is well suited for Poisson-like equations, as the SP1 -equation. The FEniCS library
14 A. JESSER ET AL.
Figure 4. Computational Model used in SPARC including the different regions of the computational
domain (left side) and the Finite Element mesh used for the reconstruction (right side). Depicted are
the moderator (green), the sample (yellow), the gamma detector (purple), the neutron detectors (red)
and on the left side the source (in blue). (a) Computational Domain and (b) Computational Mesh used
for Reconstruction.
[15,16] with the automatic mesh generator Gmsh [17] is used for the implementation. We
use PETSc [18] as linear algebra backend in conjunction with hypre [19], a library of high
performance preconditioners. For this work we use a combination of a gmres solver [20]
and an algebraic multigrid preconditioner [21]. Since the SP1 -equation is self-adjoint, a
modification of the source term is sufficient to also compute the adjoint neutron flux. The
gradient of the Lagrangian can then be computed as detailed out in Section 3.
The model used for the following simulation studies is shown in Figure 4. It depicts a
zoom into the relevant part of the computational domain. The remaining parts are filled
solely with more moderating material. It depicts the moderator (green), the sample (yel-
low), the gamma detector (purple), the neutron detectors (red) and on the left side the
source (in blue). Figure 4(a) shows the computational domain, while Figure 4(b) shows the
Finite Element Mesh used for the reconstruction including approx. 84.000 Finite Elements.
6. Numerical results
In this section, we present numerical results which show the applicability of the herein
developed theory. In Subsection 6.1, we first present a comparison of the gradient com-
puted with the Optimal Control formalism as introduced in Subsection 3.2 and a Finite
Difference approach in order to verify our implementation and the adjoint calculus in
Section 3. Thereafter, we show exemplary minimizations for which the computed gradient
is used within an L-BFGS-B scheme. In Subsection 6.2 we study both the stability of the
reconstruction under given measurement errors as well as the applicability of our method
for different materials. Throughout our simulation studies we have used fixed numerical
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 15
grids, but the investigation of the effect of measurement noise of varying intensities in Sub-
section 6.2 shows that we are not running into problems related to the so-called inverse
crime in accordance with [22].
Figure 5. Comparison of the partial derivatives of the objective function computed with Finite Differ-
ences and the Optimal Control Formalism along the line segment defined by ρ2 = 1 − ρ1 , ρ1 ∈ [0, 1].
(a) Comparison of the gradient components computed with Finite Differences and the Optimal Con-
trol Formalism and (b) Relative error of the gradient components computed with the Optimal Control
formalism.
Table 1. Average absolute errors on the reconstruction of ρ1 /ρ2 for three different error levels as well as
the average number of needed L-BFGS-B iterations.
Std. dev./total value measurement [%] Reconstruction error ρ1 [%] Reconstruction error ρ2 [%] #iterations
5 3.1 1.4 6.3
10 6.1 3.4 6.3
20 11.7 8.0 6.6
For each case, 50 simulations have been carried out. The reconstructed values are depicted in Figure 7. The used material
properties are the same as defined in Table A1 for the case Mixed Scattering/Absorption 1.
rates is 10% , while the error on the reconstructed densities is typically 5–6% . In one case
the error on ρ2 is as low as 2.6%. This seems to be an exception, however.
The regularizing effect of the term punishing deviations from the total mass introduced
in Equation (12) can be seen by comparing the upper and the lower half of Table 2. The
lower half shows the reconstruction results with the weight of the term wm set to zero.
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 17
Figure 6. The logarithmised objective function with blue, cyan and green lines showing three exem-
plary minimizations using the L-BFGS-B algorithm.
The same initial and measurement values as before have been used. Without the regular-
ization, the error of the reconstruction massively increases, from values between 2.6 and
6.1% to values between 23.3 and 66.1% . The average number of iterations also increased
considerably, from approximately five to six to values between nine and eleven. The second
test case is the sole exception. These results underline the importance of the regularization
term in order to increase the accuracy of the proposed method.
Figure 7. Reconstructed values of ρ for normally distributed measurements errors. 50 Simulations were
carried out for each error level. The average errors on the reconstruction and the number of needed
L-BFGS-B iterations can be found in Table 1.
18 A. JESSER ET AL.
Table 2. Average absolute errors on the reconstruction of ρ1 /ρ2 for four different material combinations
as well as the average number of needed L-BFGS-B iterations.
Reconstruction with regularization Reconstruction error ρ1 [%] Reconstruction error ρ2 [%] #iterations
Mixed Scattering/Absorption 1 5.4 2.6 6.3
Mixed Scattering/Absorption 2 5.2 5.5 5.5
Scattering Dominated 6.1 5.1 5.8
Absorption Dominated 5.9 6.1 4.9
reconstruction without regularization
Mixed Scattering/Absorption 1 50.0 51.8 11.0
Mixed Scattering/Absorption 2 23.7 23.3 5.8
Scattering Dominated 63.1 66.1 9.0
Absorption Dominated 46.5 48.9 9.9
The measurement values have a normally distributed error of 10%. For each material combination, 50 simulations have been
carried out. The material properties can be found in Table A1. The upper half of the table shows the reconstruction with
the term punishing deviations from the total Mass M which acts as a regularization term. The lower half of the table shows
the reconstruction without said term. For both cases, the material properties and the measurement values are identical.
where radioactive waste drums are analyzed for their elemental composition. We invert a
combined neutron-photon transport model to obtain a vector of elemental densities. In a
simplified setting considering a homogeneous sample and one-group neutron transport we
proved uniqueness of a solution. To our knowledge, this is the first mathematically rigorous
proof of uniqueness for a PGNAA problem.
To the current state of the art PGNAA spectra are evaluated in an iterative manner by
solving a fixed point problem [7,8]. This naturally reflects the dependence of the elemen-
tal composition and the neutron flux distribution in the sample. The method proposed in
this work offers an alternative. Formulating the parameter estimation problem as an opti-
mization problem we make use of a gradient descent method to solve for the elemental
composition as a parameter of the neutron transport model. As a side effect, our method
yields the elemental composition as absolute quantities in grams or kilograms. Further-
more, this approach is especially useful when several measurements are combined, since
all measurements can be combined in the objective function. For the measurements in
QUANTOM the relative formulation in [7,8] does not apply. Hence, our method offers
more flexibility to deal with a broader range of measurement tasks related to PGNAA. In
situations where both methods may be applied (standard PGNAA with one measurement
and one detector type) it is not obvious which methods yields the best performance in
terms of convergence and computation times. This will be a topic of future research.
Furthermore, we see a close connection to the theory of Diffuse Optical Tomography
(DOT). In both DOT and the PGNAA problems presented in this paper, the governing
equation is a diffusion equation. The measurement of the flux alone is not sufficient to
ensure a unique reconstruction. In DOT, it is not possible to reconstruct both the scattering
and absorption coefficients from the photon flux measurements alone [14]. Similarly, we
have seen that the neutron flux is insufficient to reconstruct the elemental densities, as
discussed in Section 4. Only the combination with the energy resolved measurements of the
gamma detector, which yields measurement results for each element individually, ensures
uniqueness. While in DOT no general uniqueness result exists due to the range of different
measurables and unknowns [25], it is still possible to obtain uniqueness results for certain
scenarios using additional assumptions on the scattering and absorption coefficients that
are to be reconstructed [26]. For PGNAA comparable studies do not exist yet since this
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 19
work represents the first result in this direction. Therefore, a complete study of possible
necessary and sufficient conditions for existence and uniqueness of solutions should be a
focus of future work.
The measurement results of the additional 3He neutron counters are included in the
objective function (Equation (11)). Monotonicity of the neutron flux with respect to the
parameter η leads to monotonicity of the counting rates and this fact was used to prove the
main uniqueness result Theorem 4.1. The proof works for any measurement device of the
quantity ‘total neutron flux’, as long as its constituting functional, representing the mea-
surement device, is linear and positive with respect to the neutron flux φ. This means, that
in praxis devices like fission chambers are also adequate to support QUANTOM measure-
ments. Furthermore, the positioning and the number of 3He detectors were not used in the
proof. Nevertheless, they might influence both the accuracy and speed of evaluations and
are worth to be further investigated in the future to optimize QUANTOM measurements.
The homogeneity of the sample is a key assumption for our proof of uniqueness in
Section 4. This case is of practical importance for PGNAA, since most measurements are
carried out with homogeneous samples, especially at research reactors [5]. Monitoring of
the total neutron flux is applied by monitor materials in this case. Hence, our results may
be adapted also for this situation.
In the case of an inhomogeneous sample our results do not guarantee a unique solution.
In the QUANTOM facility, this situation is covered by performing numerous measure-
ments at different measurement positions, corresponding to a complete surface scan of the
drum by the HPGe detector. Measurements of the total neutron flux by the 3He detectors
are also performed in every measurement position. A joint evaluation of all measurement
positions shall account for inhomogeneities in the drum. Nevertheless, the rigorous exam-
ination of uniqueness of a solution in this situation needs to be performed in the future.
It should be noted that the gradient based method in introduced in Section 3 holds for
the inhomogeneous case. Since more measurements are necessary, the objective function
will include more terms. The method relies then on the assumption of homogeneity in
discretized parts of the drum.
Acknowledgments
The authors would like to thank Framatome for the permission to use the cut-view of the QUAN-
TOM facility (Figure 1). We would also like to thank René van Geemert (Framatome) for helpful
discussions on the topic of parameter estimation problems and neutron transport. We would also
like to acknowledge the contribution of Jannick Wolters and Christopher Helmes in the development
of the SPARC code, which was used to produce the numerical results for this paper.
Disclosure statement
The authors report there are no competing interests to declare.
Funding
The work of A. J. was performed within the project RAPID, funded by the German Federal Min-
istry of Education and Research (BMBF) under the grant number 033RK094B. The work of A.J. was
also funded by the European Regional Development Fund (ERDF) project ZEBRA, funded by the
European Union and the federal state of North Rhine-Westphalia under the funding number EFRE-
0800566. The work of K.K., when affiliated with AiNT, was performed in parts within the project
20 A. JESSER ET AL.
QUANTOM, funded by the German Federal Ministry of Education and Research (BMBF) under
the grant no. 15S9406B, and in project RAPID, funded by the German Federal Ministry of Educa-
tion and Research (BMBF) under grant no. 033RK094A. The authors are solely responsible for the
content of this publication.
References
[1] Filß P. Specific activity of large-volume sources determined by a collimated external gamma
detector. Kerntechnik. 1989;54(3):198–201. doi: 10.1515/kern-1989-540326
[2] Filß P. Relation between the activity of a high-density waste drum and its gamma count rate
measured with an unshielded Ge-detector. Appl Radiat Isot. 1995;48(8):805–812.
[3] Coquard L, Havenith A, Kettler J, et al. Non-destructive material characterization of radioactive
waste packages with QUANTOM. In: WM2020 Conference; 2020 March; Phoenix, Arizona.
[4] Coquard L, Hummel J, Nordhardt G, et al. Non-destructive verification of materials in waste
packages using QUANTOM. EPJ Nuclear Sci Technol. 2023;9(55 doi: 10.1051/epjn/2022043
[5] Molnar G. Handbook of prompt gamma activation analysis with neutron beams. New York:
Springer; 2004. doi: 10.1007/978-0-387-23359-8
[6] Jesser A, Krycki K, Frank M. Partial Cross-Section calculations for PGNAA based
on a deterministic neutron transport solver. Nucl Technol. 2022;208(7):1114–1123. doi:
10.1080/00295450.2021.2016018
[7] Holloway J, Akkurt H. The fixed point formulation for large sample PGNAA – part 1: theory.
Nucl Instrum Methods Phys Res A. 2004;522:529–544. doi: 10.1016/j.nima.2003.11.401
[8] Akkurt H, Holloway J, Smith L. The fixed point formulation for large sample PGNAA – part
2: experimental demonstration. Nucl Instrum Methods Phys Res A. 2004;522:545–557. doi:
10.1016/j.nima.2003.11.400
[9] Cercignani C. The boltzmann equation and its applications. Springer New York; 1988. doi:
10.1007/978-1-4612-1039-9
[10] Rising ME, Armstrong JC, Bolding SR, et al. MCNP R
Code Version 6.3.0 Release Notes. Los
Alamos (NM): Los Alamos National Laboratory; 2023. LA-UR-22-33103, Rev. 1.
[11] Gilmore G. Practical Gamma-Ray spectrometry. Wiley; 2008. doi: 10.1002/9780470861981
[12] Seltzer S. Tables of X-Ray mass attenuation coefficients and mass energy-absorption coeffi-
cients, NIST Standard Reference Database 126; 1995.
[13] Hinze M, Pinnau R, Ulbrich M, et al. Optimization with PDE constraints. Dordrecht: Springer;
2009. doi: 10.1007/978-1-4020-8839-1
[14] Arridge S, Lionheart W. Nonuniqueness in diffusion-based optical tomography. Opt Lett.
1998;23(11):882–884. doi: 10.1364/OL.23.000882
[15] Alnæs M, Blechta J, Hake J, et al. The FEniCS project version 1.5. Archive Numer Softw.
2015;3(100):9–23.
[16] Logg A, Mardal K, Wells G. Automated solution of differential equations by the finite element
method. Springer; 2012. doi: 10.1007/978-3-642-23099-8
[17] Geuzaine C, Remacle F. Gmsh: a three-dimensional finite element mesh generator with built-
in pre- and post-processing facilities. Int J Numer Methods Eng. 2009;79(11):1309–1331. doi:
10.1002/nme.v79:11
[18] Balay S, Abhyankar S, Adams M, et al. PETSc users manual. Argonne National Laboratory;
2020. ANL-95/11 – Revision 3.14.
[19] Falgout R, Meier Yang U. hypre: a library of high performance preconditioners. In: Interna-
tional Conference on Computational Science; 2002 April 21–24; Amsterdam, The Netherlands:
Springer.
[20] Saad Y, Schultz M. GMRES: a generalized minimal residual algorithm for solving nonsymmet-
ric linear systems. Siam J Sci Stat Comput. 1986;7:856–869. doi: 10.1137/0907058
[21] Ruge J, Stüben K. Algebraic multigrid. In: McCormick S, editor. Multigrid methods, volume 3
of Frontiers in applied mathematics. SIAM; 1987. p. 73–130. doi: 10.1137/1.9781611971057
[22] Wirgin A. The inverse crime; 2004. doi: 10.48550/arXiv.math-ph/0401050
APPLIED MATHEMATICS IN SCIENCE AND ENGINEERING 21
[23] Zhu C, Byrd RH, Lu P, et al. Algorithm 778: L-BFGS-B: fortran subroutines for large-
scale bound-constrained optimization. ACM Trans Math Softw. 1997;23:550–560. doi:
10.1145/279232.279236
[24] Virtanen P, Gommers R, Oliphant TE, et al. SciPy 1.0: fundamental algorithms for scientific
computing in python. Nat Methods. 2020;17:261–272. doi: 10.1038/s41592-019-0686-2
[25] Gibson AP, Hebden JC, Arridge SR. Recent advances in diffuse optical imaging. Phys Med Bio.
2005 Feb;50(4):R1. R43doi: 10.1088/0031-9155/50/4/R01
[26] Harrach B. On uniqueness in diffuse optical tomography. Inverse Probl. 2009 Apr;25(5):055010.
doi: 10.1088/0266-5611/25/5/055010
[27] Gilbarg D, Trudinger N. Elliptic partial differential equations of second order. Berlin: Springer;
2001. (Classics in mathematics).
Proof: We carry out the proof by constructing a contradiction. Let us first define u = −λ so that
Equation (29) can be rewritten as
L(x)u(x) = Q(x) in D. (A1)
with the differential operator defined by
L(x)λ(x) = ( − η(x))λ(x). (A2)
The
coefficients of L are bounded and the partial differential equation is strictly elliptic. We also have
−ηv dx ≤ 0 for v(x) ≤ 0 since η > 0 and Lu ≥ 0 since Q is non-negative. Therefore we can apply
the strong maximum principle (theorem 8.19 in [27]).
If supD u ≥ 0, then according to the strong maximum principle u is constant in D and for u = 0
we have
ηv dx = 0 ∀ v ≥ 0, v ∈ C01 (D). (A3)
D
While u = 0 cannot satisfy Equation (A1) since Q > 0 on a non-null subset of D, u = 0 cannot
satisfy Equation (A3) since η > 0 for non-null subset of D. Therefore, we have supD u < 0 and thus
inf D λ > 0.
Proof: Let u1 and u2 be solutions of Equation (30) and v = u1 − u2 . For w = v∇v, the divergence
theorem
∇ · w dx = w · n ds (A4)
D ∂D
22 A. JESSER ET AL.
yields
∂v
(∇v)2 + vv dx = v
ds. (A5)
D ∂n
∂D
Using −v + αv = 0 in D and inserting the boundary condition, we obtain
2
∂v
(∇v)2 + αv2 dx = −β ds. (A6)
D ∂D ∂n
Since the integral on the left side is non-negative and the integral on the right side non-positive,
∂v
both integrands are identically zero. Therefore, ∂n = 0 and from the boundary condition also v is
identically zero on ∂D. From the weak maximum principle [27] follows that v is identically zero
on D.