LIFE CYCLE COST ANALISYS APPLIED TO

REINFORCEMENT CORROSION REPAIR

D. Izquierdo1 Dr. Civil Engineer

ABSTRACT.

Reinforcement corrosion has been shown to be the most important cause of

structure disruption, and therefore the most extensive cause of repair in concrete
structures. Nowadays, the procedures for repairing reinforcement corrosion are
more or less well known and if certain types of recommendations are followed, the
repair procedures work properly. One of the procedures for selecting the most
appropriate repair method is the economical optimisation of the whole cost of the
repair procedure, or the Whole Life Cycle Cost (LCC). Present paper describes
several models and values for accounting the LCC of a structure and their
application to reinforcement corrosion repair

RESUME.

La corrosion des armatures est considérée comme la principale cause de

dégradation des structures en béton et, par conséquent, ce phénomème conduit á
un grand besoin de réparation. Actuellement, les processus de réparation de la
corrosion des armatures sont plus ou moins connus et, si queques
recommendations sont suivies alors les réparations seront durables. La plus
appropiée des méthodes de réparation est basée sur l´optimisation économique du
coût total. Elle est connue comme l´analyse du Cycle de Vie Totale (LCC). La
présente communication décrit quelques modèles mathématiques utilisés pour
réaliser un LCC en l´appliquant spécifiquement á la réparation des structures en
béton armé.

Keywords: Life Cycle Cost, Probabilistic Approach, Discount Rate,

Reinforcement Corrosion, Time Dependent Reliability

1. D. Izquierdo. Instituto Eduardo Torroja – CSIC, Serrano Galvache s/n,

28033 Madrid (Spain) izquierdo@ietcc.csic.es

Do not add page numbers to the final version

 INTRODUCTION

Life Cycle Cost Analysis is not a new technique for optimisation civil
infrastructure 1, 2. Although there are new advances in research about optimisation
in the LCCA 3, 4 the main purpose of traditional procedures is the whole concept of
infrastructure design, which includes all phases of the infrastructure’s life:
Building, service life, repair and restoration, residual service life and disposal.
However, the repair stage can be considered as a new project phase and therefore
new alternatives are presented and an economical optimisation is suitable to be
employed. Of course the economical optimisation must comply with al the
requirements proposed by owners and other infrastructure agents.

LCCA is only a whole optimisation of the money used in the infrastructure. Of

course, the most easy expression for accounting this money can be obtained by
adding the building cost to the maintenance costs and the cost of a possible
infrastructure failure (equation 1):

CT = C I + CM + C f Equation 1

For accounting this equation it is essential to assess money in equivalent currency

by means an appropriate discount rate and Equation 2, where NPV is the Net
Present Value of a nominal cost Cn, r is the discount rate, and t is the actualisation
time of the cost.

1
NPV = Cn Equation 2
(1 + r )t
One of the most interesting new applications of LCCA is the selection of an
appropriate repair procedure in structures. Usually the selection of a repair
procedure or system is performed taking into account the actual investment in the
structure (the Direct Cost of the repair), thus the LCCA allows to evaluate not only
the effect of the repair cost, but also the expected service life of the repair and
therefore new possible costs of a failure in the repair.

Of course, it exists many types of LCCA depending of the amount of data and
information about the performance we have. The starting point should be a rational
classification of the typical LCCA used.

TYPES OF LIFE CICLE COST ANALISYS

A rational classification of LCCA can be established taking into account the

uncertainty of the input data, thus it can be possible to speak about:

- Deterministic Life Cycle Cost Analysis (DLCCA).

- Probabilistic Life Cycle Cost Analysis (PLCCA).

Within the PLCCA, the costs can be accounted by deterministic quantities or by

probabilistic quantities. Of course, the amount and quality of information is higher

Do not add page numbers to the final version

as higher is the level of the LCCA however the amount of input data should be
increased in an equivalent manner.

DETERMINISTIC LIFE CYCLE COST ANALYSIS

It is the simplest form of LCCA. It only considers deterministic costs and expected
service life and durability. It can be consider as a first step in the selection
procedure for repair strategies. Although their simplicity, usually practical
approaches of economical optimisation of a DLCCA.

The expression that can accounts for the total expected cost of the structure, taking
into account the maintenance and repair interventions is collected in Equation 3,
where each expected cost should be translated into actual currency by means the
discount rate r.

n
CM ,i n C f ,i
CT = CI + ∑ +∑ Equation 3
i =T 1 (1 + r ) i
i =T 1 (1 + r )i
The parameters in Eq. 3 are:

- CT total cost of structure.

- CM,i Maintenance cost at time i.
- Cf,i failure costs at time i.
- r discount rate

The distribution of cost, should be considered as localised – in – time costs,

because the service life in DLCCA is deterministic. An example is shown in
Figure 1.

Figure 1. Example of distribution of costs in DLCCA

Distribution of Costs in time

120
Maintenance
100 Repair
80 Strengthening
Cost

60

40
20

0
0

10

20

30

40

50

60

70

80

90

0
10

Time

PROBABILISTIC LIFE CYCLE ANALYSIS

It can be considered an evolution of DLCCA where all variables (except costs) are
considered as stochastic, therefore the costs of maintenance, repair and
strengthening are related to their probability of occurrence and their PDF. Thus,

Do not add page numbers to the final version

another type of costs should be considered. It is the cost of structure collapse,
whose probabilities are certainly low to be taken into account in the DLCCA.

A more general expression that takes into account the probabilistic approach in the
evaluation of LCC, can be rewritten in Equation 4, where it should be noticed that
in the case of costs due to Maintenance (CM), Repair / Strengthening (CR,S) or
Collapse (Cf) the sum is extended from the time t=1 because the probability of
repair/maintenance or structural failure is higher than 0.

n
CM ,i
CT = C I + ∑ PF ,M (ti ) +
i =1 (1 + r )i
Equation 4
n
C R ,S ,i n C f ,i
∑P (ti ) + PF (ti )
i =T 1
F ,R ,S
(1 + r )i ∑
i =1 (1 + r )i
Thus, the distribution costs throughout time, are not localised as in DLCCA if not
continuous because the cost is multiplied by the probability which usually has a
continuous distribution with time. In this sense, the product of probability by costs
is usually known as damage.

MODELS FOR REPAIR / MAINTENANCE COSTS

One of the most difficult points of using this type of approach (either
Deterministic or Probabilistic) is the determination of realistic models for taking
into account the costs due to repair / maintenance. The word model describe a
mathematical expression that relate expected costs due to Repair / Maintenance
and the geometrical, disposition or localisation characteristics of the structure.
Thus, a relatively short expression that can describe the costs due to a Repair /
Maintenance operation in a structure can be written as shown in Eq. 5.

C M / R = C D + C H + CU + C E Equation 5

In Eq. 5 the costs included are: Direct cost of repair (CD), Cost of loss of contents
or fatality and injury loses (CH), Cost of users (CU) and Socio – Economic loses
(CE). In each case the cost can be multiplied by their appropriate Probability
Density Function (PDF) and use PLCCA.

It can be considered that assess a realistic (more or less) appropriate value of

direct const can be carried out. Typical prices list and standardisation cost can be
directly used. However the evaluation of indirect costs is quite difficult and
depends on many factors. Some notes and examples will be shown in advance.

- Indirect costs models for buildings

Indirect cost models are quite different for building and for bridges. Typical
models for buildings include costs as shown in Eq. 65 where are included cost of
contents (CC), economic loss due to business interruption (CB), cost of injury in
repair processes (CIn) and costs of possible fatality in the repair process (CFa):

Do not add page numbers to the final version

 C M / R = C D + CC + C B + C In + C Fa Equation 6

Cost of contents can be easily obtained form common price lists and cost of
business interruption can be computed by accounting past economical reports of
the factory or the cost of new rent if the building is dedicated to services, offices
or residential. It is usually presented in terms of initial construction cost of the
structure.

Cost of CIn and CFa are considerably difficult to analyse, because the special
influence of human life. Although in old reports and documents it was said that
trying to evaluate the economical value of a human life is rather inmoral, new
approaches suggest that the Life Quality Index may be assessed using LCCA6 .

- Indirect costs models for bridges and roads

In general roads and bridges costs consist of five major cost item: vehicle
operating costs, time delay costs, safety and accident costs, comfort and
convenience costs and environmental costs. It is usually considered that two first
are the most relevant although in order to evaluate them factors such as traffic
network, location of the infrastructure and information of rehabilitations must be
considered. A expression for accounting time delay costs can be written in Eq. 7.

J
 N

CTD = ∑ nPj ·T j ·u1 j 1 − ∑ ri ∆t +
j =1  i =1 
N J
Equation 7
∑ ·∑ r ·n
i =1 j =1
i Pj ·T j ·u1 j + rij ·nPij ·Tij ·u1ij ∆tij

where: j is an index representing the type of vehicle (business and non – business,
etc). nPj is the average number of passengers in vehicles type j, Tj is the Average
Daily Traffic Vehicle of type j, u1j is the unitary cost of unit of time for operator j,
ri is the rate of occurrence of route i and ∆t is the increment in time of route i.

Costs of vehicle operation can be computed using Eq. 8 where it is considered the
wages for each type of vehicle and the cost of fuel.

 
∑ (T 0 j ·u 2 j ) 1 −
J N
C TD =
j =1 
∑ r ∆ t +
i =1
i

∑ r ·∑ (T ·u 2 j )∆ t i +
N J

i 0j Equation 8
i =1 j =1

ri ·∑ T 0 j ·(u 3 ij l di − u 4 ij l d 0 ) + T ij u 2 j ∆ t i
N J

∑ i =1 j =1

Where u2ij is the average operator wages for each type of vehicle, u3ij is the
average unit fuel cost pert unit of length on each detour route, uiij is the average
unit fuel cost per unit of length in the original route, li and l0 are the length of the
detour and original route, respectively.

Do not add page numbers to the final version

For other costs such as socio – economic costs, it is necessary to evaluate the Input
– output table for the productivity of the country.

IMPLICATIONS FOR TIME – VARIANT PLCCA

From a direct examination of Eq. 3, where the Probabilistic Life Cycle Cost
Analysis is described, it can be seen that each cost must be actualised in time. This
actualisation, as was explained above should be done continuously because the
implication of probabilities in the computation. To compute directly expressions
such as Eq. 3 is considerably difficult due to the calculation of probabilities with
time. One of the simplification proposed by authors7 is to compute Eq. 3 in a
similar manner as Eq. 2 at the end of the expected service life of the repair. One
possibility for simplification is to compute costs at a deterministic point in time
and actualise currency values in the same manner.

n
C M ,i tL
C M ,i CM ,i
∑ PF ,M (ti )
i =1 (1 + r ) i
= ∑ PF ,M (ti )
i =0 (1 + r ) t act
=
(1 + r )t act
PF ,M (t L ) Equation 9

The time – factor tact takes into account the effect of increasing evolution of
degradation with time. Of course tact depends essentially of two factors: PDF of
service life and discount rate. Figures 1 and 2 show the net value of the damage
cost as a function of the mean service life for a Log – normal distribution and
100% of variation.
Figure 2 and 3. Example of Damage costs and tact

0.6 180

160
0.5
140

0.4 120

100
Damage

T_act

0.3
80

0.2 60

40
0.1 20

0
0
0 50 100 150 200 250 300
0 50 100 150 200 250 300
Mean Service Life
Mean Service Life

r=2% r=4% r=8% r=12% r=2% r=4% r=8% r=12%

REALISTIC VALUES FOR DISCOUNT RATE

It can be easily seen that the results are quite different depending of the discount
rate (r) we chose for analysis. This aspect is one of the principal inconveniences
of LCCA. Of course it is known by politicians as a way to force or delay future
inversions in infrastructure or whatever type of business. Thus, it is commonly
accepted that national administrations force the technicians to use specific r values
depending on the type of inversion, these values are varying from 12% to 6%1. An
economical definition of r can be obtained as the difference between rate of

Do not add page numbers to the final version

interest and inflation, as shown in Eq. 10. Application of Eq. 10 to statistical
economical data of several countries has shown that it is unrealistic to adopt
values of discount rate above 2 – 5%. Figure 4 shows the values obtained for
several countries, taking into account time life windows of 50 and 10 years.
 1+ i 
r =  −1 Equation 10
1+ a 
Figure 4 . Realistic Discount Rate in Europe
7 6.43

6
5.1
Actualisation rate [%]

4 3.51 3.59
2.91
3
2.02 2.01 2.07
1.81
2 1.5
1.25
1

2
in
in

ax

ax
0
ax

50
0

en
M

M
-5

M
-5

M
M

n-

ed
m

0
SA

0
0

ai

-2

-2
10
0
er
0

-1
-1

Sw
-1
-1
U

Sp

EU

EU
G

n-
m
SA

m
SA

er

ai
er
U

Sp
G
U

CONCLUSIONS
Life Cycle Cost Analysis has been developed many years ago, and their
application to building projects is being increased in last twenty years. One of
their possible applications is the selection of adequate repair / strengthening
methods to be employed in infrastructure. The models for analysing each cost are
presented and identified, one of the problems is to adopt numerical values for each
variable.
In addition, repair / strengthening activities are linked to a degradation process
that involves the time effect. The time adds an additional difficult in the
computation process and more uncertainties. One possibility is to translate
calculations to DLCCA by means the actualisation time as was presented, making
considerably easy the results.
Finally, all results obtained in LCCA, are strongly dependent of discount rate
adopted in the calculations. A specific study is presented showing that realistic
values are located in the range of 2 – 5%.

REFERENCES
1. Tilly, G. P. , Principles of Whole Life Costing Safety of Bridges, 1997, Ed. By Parag C.
Das, Pag 138 – 144.
2. Ferry, D.J.O. Flanagan, R. Life Cycle Costing – a radical approach. CIRIA report, nº 122
3. Cho, H.N. Life Cycle Cost Effectiveness for Design and rehabilitation of Civil
Infrastructures Proceedings of SEWC’02.
4. Cho, H.N. Development of Bridge Life – Cycle Cost Analysis Model Final report to
Korea Infrastructure Safety and Technology corporation 2002.
5. Ang A, De Leon D. Determination of optimal target reliabilities for design and upgrading
of structures. Structural Safety,(1997) 19(1).Pag. 91 – 103.
6. Rackwitz R. Optimisation – The Basis of code making and Reliability Verification.
(2000) Structural Safety , 22 Pag 27 – 60.
7. Siemes. A. J. M. Vrouwenvelder A. C. Van del Beukel A. (1985) Durability of buildings:
a reliability analysis. HERON (1985) Vol. 30 nº 3 Delft University of Technology.

Do not add page numbers to the final version