United States
                     Environmental Protection
                     Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
                     Research and Development
EPA-600/S3-84-012  Mar. 1984
oEPA          Project  Summary

                     Damage  Cost  Models  for
                     Pollution  Effects on  Material
                     Edward F. McCarthy, Alexander R. Stankunas,
                     John E. Yocom, and Douglas Rae
                       Two  damage cost models were
                     developed to quantify the effects of
                     ambient air pollutants on manmade
                     materials exposed in urban environ-
                     ments. The models use existing physical
                     damage functions, estimates of material
                     in place and average repair or replace-
                     ment costs  to  calculate the use life
                     maintenance costs as a function of air
                     pollutant concentration.
                       The first model, called the "prevailing
                     practice model", assumes that existing
                     maintenance practices represent a
                     rational response to current levels of
                     pollution and  material properties.
                     Information  on the frequency of main-
                     tenance actions and the rate of damage
                     predicted by existing physical damage
                     functions is  used to derive a "critical
                     damage level". This  critical damage
                     level, defined as the amount of damage
                     that is usually accepted before remedial
                     action  is taken, is assumed to be
                     constant. The change in  the rate of
                     damage with changes in pollutant
                     concentration can then be used to
                     calculate a  change in maintenance
                     schedule, which, in turn, is converted to
                     a change in maintenance costs over the
                     use life of the material system under
                     study.
                       The second model, called the "least
                     cost model" does not make any assump-
                     tions about the appropriateness of
                     current maintenance practices. Instead,
                     the critical damage levels and mainte-
                     nance criteria are directly specified by
                     the user. The model then calculates the
                     system use life  costs for maintenance
                     schedules based on these criteria for
                     different pollution levels.
                       Each  model has its advantages and
                     disadvantages. The prevailing practice
                     model  is easy  to  use but  is highly
dependent on the accuracy of informa-
tion on existing maintenance practices
and the appropriateness of the assump-
tion that such practices represent the
most  rational  response  to existing
conditions.  The least cost model is
more versatile in that it can be applied to
conditions that are not representative
of the existing situation, but it requires
more detailed  input and assumptions.
Both models are highly dependent on
the accuracy  of existing physical
damage functions.
  This report presents both approaches
and demonstrates their application to
calculating the cost of sulfur  dioxide
damage to steel, zinc and paint, total
suspended particulate matter damage
(soiling) of clean surfaces, and ozone
damage to elastomers.

  This Project Summary was developed
by  EPA's Environmental Sciences
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project  Report ordering
information at back).

Introduction
  The physical effects of air pollution on
manmade materials  have long  been
recognized and a considerable amount of
knowledge on physical effects has been
accumulated. Economic estimates of the
cost of damage have been made based on
this knowledge, but there is little confidence
in these estimates because of questions
concerning the accuracy of key input data
and the lack of sophistication of the
techniques used. The importance of
accuracy in making economic assessments
for  cost-benefit comparisons and con-

-------
sideration of secondary  air quality
standards is obvious.
  The  damage  costs due to  an  air
pollutant must be calculated in  light of
natural  limits to the useful life  of  the
material affected and options for repair,
replacement or  substitution. Lifetime is
distinguished from  the  use life  of  the
system  as being the time  of exposure
experienced until a critical damage level
is reached at which action is taken
and/or real costs are incurred. Use life is
defined as the time period over which the
material system is expected  or needed to
be used. For example, children's shoes
usually have a lifetime significantly
longer than their typical use life. That is,
they are usually discarded  long  before
they have physically deteriorated to  the
point at which  they can no longer be
worn. Therefore, little attention is paid to
repairing  children's shoes. A painted
house,  on the other hand, has a use life
far in excess of the  lifetime  of the paint.
Accordingly, maintenance of the house
paint is  usually performed.
  The  incremental cost of pollution
damage, that is, the cost associated with
damage above that expected from normal
wear or use in an  unpolluted ("clean")
environment, is the most important
factor in the consideration  of the costs
and  benefits of pollution control. More
specifically, the damage to a  material by a
given pollutant  must be assessed if the
benefit  to  be gained  by reducing that
pollutant is to be calculated.
  Accurate estimates of the  disbenefit of
air pollution damage to  materials, (con-
versely, the benefits  of  avoiding such
damage),  are difficult  to derive. Such
estimates must not only take into account
many important physical and chemical
interactions, but  must  also consider
socioeconomic  factors. Aesthetic judg-
ment, an  awareness of alternatives  and
the cost of capital  are only  a few of the
factors  which  can strongly influence
incurred costs.
  In  order to determine the costs associ-
ated with the exposure  of  materials to
ambient pollutant levels, several types of
information are  necessary. This includes
the  distribution and  exposure  of  the
materials  of interest, the rate at which
damage is incurred under actual exposure
conditions, the amount of damage which
necessitates remedial  or preventive
action (critical damage level), and finally,
the type and cost of remedial  or preventive
action actually taken.
  The  objective of this work  was to
develop a relatively simple method for
estimating the damage  cost  of  air
pollution with regard to its effects on
nonliving materials, and to establish a
framework  by which the estimates may
be improved when additional information
becomes available. To meet this goal two
methods for estimating the damage cost
of air pollution were developed. The first
method is a simple approach in which
current maintenance practices, determined
from surveys, are coupled with  existing
pollutant levels and  physical damage
functions  and are  extrapolated into
general rules  or functions. These rules or
functions can  then be applied in hypothet-
ical situations where factors such as the
pollutant level are varied over a limited
range  and the resulting  change  in
maintenance  and replacement costs can
be calculated.
  The second method is  a more complex
model  which  uses  available pollutant
damage  functions and  specific critical
damage levels to calculate damage costs
directly. The complex model is somewhat
more theoretical but it allows considera-
tion of alternative maintenance strategies
that  may be significantly different from
those currently influencing  actual prac-
tice.  The complex model can thus  be
used to derive  an estimate of costs for
conditions or  maintenance strategies not
currently in use, including a theoretical
ideal or least  economic cost approach.

Damage  Cost Model

General Concepts
  There are two fundamental approaches
to estimating  the economic impact of air
pollutants on  materials. The first Is a
direct,  empirical  comparison of total
expenditures  and/or a loss of amenity
due to  materials damage under different
atmospheric pollutant conditions, followed
by the direct development of quantitative
relationships between cost and pollution.
  The second approach  is based on the
calculation  of the  physical damage from
a given atmospheric pollutant concen-
tration  by  means of  physical damage
functions and the quantification of the
economic and aesthetic responses to that
damage, through economic damage
functions.  The cost versus pollution
relationships are thus derived from
calculated  effects,  and  are not based
solely  on observation.  Although less
direct,  this  analytical approach  has the
advantage  that it is  less sensitive  to
common sources of error  such as regional
differences in climate, population, income,
mix of materials and spurious correlation,
than the comparative approach.
  Two  models employing  the  second
approach have been developed. The first
model,  called the  prevailing practice
model, reflects current maintenance and
replacement practices. It  is assumed
in this model that such prevailing practice
is the result of well-informed decisions
and represents the best possible response
to local conditions. The second model, or
least cost model, uses pollutant  damage
functions,  current ambient pollutant
concentrations, specific  critical  damage
levels and a maintenance or replacement
decision matrix to determine a least cost
economic strategy  independent of  the
prevailing practice.
Prevailing Practice Model
  The prevailing practice model  is based
on several  key assumptions. First, it is
assumed that  the current strategies for
the use of material  systems incorporate
decisions based on accurate information
of both  physical and socioeconomic
factors.  This  assumption  may  be  of
limited validity due to the rapid introduc-
tion of new materials and the relatively
recent changes  in  ambient air quality
which have drastically changed material/
environment interactions. Strategies that
are used today may not yet reflect these
changes since the consequences of the
changes are not yet apparent. This simple
model  is still a  useful tool, however,
because it automatically  includes a
number of variables which are difficult, if
not impossible, to define precisely.
Several of these variables are socioeco-
nomic as well as physical in nature, such
as the level of damage which  prompts
repair and replacement. It is also a model
which  may  be applied with reasonable
accuracy and without  extensive (and
expensive) data gathering.
  A schematic outline  of the prevailing
practice model approach is presented in
Figure 1. The prevailing practice model
requires information on the lifetime of
materials systems in the ambient environ-
ment which can usually be derived from a
survey of currently practiced maintenance
strategies.  Physical  damage functions,
which relate the concentration of a given
pollutant to the rate of physical damage
are then selected. The  current pollutant
level and other environmental factors are
then  used in  the physical damage
function to define a  critical damage level
that is  consistent  with the prevailing
lifetime  of  the  material. The  damage
function and defined critical damage level
are then applied  with  the projected
pollutant levels of interest and a projected
lifetime of the material  is calculated. The
projected lifetimes are then  used to
determine  the costs attributable to
pollutant damage at each level of pollution.

-------
           Determine Prevailing
          Maintenance Strategy
         Use Damage Function for
          Pollutant and Material
           to Derive Perceived
        Critical Damage Level (CDL)
      Use Perceived CDL to Calculate
       Maintenance Frequency in a
       Pollutant Free Environment
      Use Perceived CDL to Calculate
         Maintenance Frequency
           at Various Levels of
               Pollutant
         Use Simple Annualized
        Cost Approach to Calculate
        Total Costs and Incremental
          Costs Due' to Pollutant
Figure 1.    Prevailing practice model.


Least Cost Model
  The  least cost model  is much  more
complex than  the prevailing  practice
model  described above. Instead of simply
accepting the  typical period between
maintenance activities as the lifetime of
the  material,  this model is  used to
calculate the  maintenance schedule
most appropriate for minimizing total
cost. These calculations are based on
physical  damage functions,  externally
derived  critical damage levels  and
information on economic factors such as
the cost  of capital (discount  rate) as  a
function of income for various groups.
  As  in  the  simple  prevailing  practice
model, the  first  step is to define the
nature of  the materials  system  and
pollutant interactions of interest. However,
unlike  the simple model, the progressive
changes  in  the rate at  which  both
physical and economic damage accumu-
lates as  a function of previous history
must be  included. The least cost model
accounts for these changes by  allowing
the definition of the material subsystem
to change as described by critical damage
levels.  The rate of accumulation of both
physical  and economic damage is  also
changed as the definition of the  material
                                       Input:
                                         SOz Concentrations
                                         Maintenance Costs
                                         Interest Rates
                                         Material in Place
                                              I
                                        Define Maintenance
                                          Strategies and
                                          Critical Damage
                                           Levels (CDL)
                                                                                       i
Replacement


Calculate t (years)
Using CDL and Physical
Damage Functions
Compare t to Use Life
U


NO
Maintenance
Required

          Policy of No
          Maintenance
                                                                       No
     Preventive
Maintenance Performed
       Calculate t Using CDL
       and Physical Damage
           Functions
              I
       Compare t to Use Life
                                                                                         Yes
 Calculate Frequency of
 Selected Maintenance
                                       Calculate Net Present
                                              Value
Figure 2.    Least cost model flow diagram.

 subsystem  is varied. Since critical
 damage levels are  usually  defined  in
 terms of either changes  in the rate  of
 accumulation of  economic damage  or
 changes in system utility, the method is
 not as complex as it  initially appears.
   In practice the critical damage levels
 are specified in the model. The appropriate
 physical damage  functions are used  to
 define the length of  time required  to
 reach  each critical  damage  level  for a
 given set of environmental conditions. A
 cost factor is  associated with the mainte-
 nance  performed  within the period
 between changes  in critical damage
 level. The model then calculates the use
   life  costs for different maintenance
   schedules and adjusts the costs to
   account for the cost of capital.
     The  impact  of a  change in pollutant
   level is determined by the total minimized
   costs of maintenance at different levels of
   pollutant concentration.
     A schematic diagram of the least cost
   practice model is illustrated in Figure 2.

   Model Application

   Prevailing Practice
     The prevailing practice model was used
   to determine incremental damage costs
   associated with  changes  in  annual

-------
average SOa and TSP concentrations for
bare galvanized steel and painted steel
exposed  in  an urban  environment.
Normal  maintenence  practices for the
repair and replacement of these materials
were determined in a limited local survey
of the Boston  area. The maintenance
practices  reviewed include  those  for
highway stuctures (bridges, signs, poles,
guardrails), chain-link fencing, and electric
transmission towers.  One  interesting
result of the survey was the discovery
that the majority of galvanized materials
are not routinely  maintained. Only a
relatively  small percentage of the total
stock  of bare galvanized  steel products
has deteriorated to the point  where
maintenance has been required, although
much of it has been removed for other
reasons (aesthetic appearance, damage by
impact or physical wear).
  Accordingly, although damage to struc-
tures  such  as  chain  link fencing and
electric transmission  towers could be
significant, there is not enough current
maintenance history to judge the degree of
significance through the prevailing practice
model.
  In another application,  the amount of
painted surface in an  urban area was
estimated, the  prevailing maintenance
practice  was  determined  through a
survey  of residential house painting
practices  in  the  Boston area and  the
incremental damage costs attributable to
annual average pollutants was calculated.
Application  of this  model with the
existing mathemetical  damage function
for paint indicates that only  a small
portion  of  the costs of painting is
attributable to S02 damage.
  The  evaluation of the  damage costs
associated with total suspended particu-
late matter (TSP) and soiling  was limited
by the available pollutant-material damage
functions  and a lack of data on  normal
practice and critical damage levels.  For
soiling the potential disbenefits  are
primarily  in the form  of  aesthetic costs
rather than  direct costs associated with
changes in  maintenance practices. The
economic impact of soiling, therefore,
could not be determined accurately with
the prevailing practice  model.
  The relationship between  ozone and
damage to rubber tires  was analyzed and
two areas of disbenefit were identified.
On a regional basis, reduction in ozone
concentration would result in small
benefits to  the retread  industry in the
form of an increased number of available
casings. Benefits from reduction in ozone
on a  national level could occur as the
amount of antiozonant added to tires by
manufacturers was reduced to provide
the same level of protection currently
afforded, resulting in lower costs. How-
ever the benefits associated with changes
at either the national or regional level
were calculated to be relatively small.
  The results of the prevailing practice
model applications are presented in Table
1.

Least Cost
  The least cost model combines physical
damage functions and an economic
approach using  the net present value
technique (NPV) to predict the least cost
approach  to  addressing damage to
materials exposed outdoors. A variety of
maintenance  strategies (including no
maintenance); the cost  associated with
each maintenance strategy; and various
ambient concentrations  of air pollutants
were considered.  Other  factors included
in the analysis were:
   • The use life of the material, defined
     as the period of time that the system
     of  which the material is a part is
     expected to  perform a  particular
     function.
   • The value of the material  in place
     reflected by the cost of total repair or
     replacement.
   • The critical damage levels, defined
     as  the amount of damage  which
     prompts a decision  for maintenance
     or replacement of the material.
   • The interest rates for use in calculat-
     ing the net present value.
  The least cost model was configured to
calculate the damage costs associated
with various annual average SOzconcen-
trations  (0, 20, 40, 60,  80, 100/yg/m3)
and interest rates  of 5 percent, 10 percent
and 15 percent.
  An example of  the results of the least
cost model are presented in Table 2. In
this application, damage to painted wood
residential  structures from exposure to
ambient SOz was evaluated using three
maintenance strategies. The model was
also applied to study bare galvanized steel
    chain  link fence and  painted metal
    reactions with ambient SOa-
      The  most  dramatic  result of  the
    analysis presented  in  Table  2 is  the
    importance of the discount rate used to
    calculate net present value costs. High
    discount rates put a heavy emphasis on
    near term expenditures, and this fact is
    reflected in the relatively small  influence
    of both maintenance  strategies  and
    pollutant levels on calculated maintenance
    costs. The time-to-first maintenance has a
    powerful  effect  on  total costs. Since
    damage by pollutants to most  materials
    used  in permanent  construction takes
    several years to become evident, the use
    of the  net present value method of  cost
    accounting tends to reduce substantially
    the apparent  economic impact.

    Conclusions
      The  development of these two damage
    cost models demonstrated that there are
    several key areas  where additional
    information is essential  to develop more
    accurate estimates. The basic  economic
    approaches of using prevailing practice
    and least  cost analyses  to develop costs
    are basically  sound. However, uncertain-
    ties as to the  amounts of exposed
    material in place on  a nationwide scale;
    the lack of knowledge of the response of
    the industrial and private sectors to the
    effects of  air pollution; and the failure of
    several of the physical damage functions
    together to  address the  major factors
    leading to maintenance actions, undermine
    the  accuracy of economic estimates
    made  using these models.
      Additional information  must be obtained
    on normal or prevailing practice to
    determine  how  the  public  and  private
    sectors respond  to conditions prompting
    maintenance and replacement and the
    estimation techniques for quantifying the
    amount of susceptible materials in place
    must be improved. These data would be
    used as primary input to  the estimation of
    material damage nationwide. The damage
    functions for steel and zinc in  the
 Table 1.    Example of Applications of the Prevailing Practice Model

                                   	Pollutant Concentration (pg/m3)
    Pollution/Material	
Incremental Per Capita Annual Cost (1981$)
 SOi/Bare Galvinized Steel
 TSP/Clean Surfaces
                NA"
                NA"
SOi/Painted Wood
Oz/ Rubber (national)
0
0
40
.14
20
9.07*
60
.36
40
3.07*
80
.54
100
28.49
100
.59
 " Insufficient maintenance history to apply prevailing practice model.
 b Inadequate damage functions to apply prevailing practice model.
 c Cost calculated for changes in lifetime of integer years.

-------
Table 2.    An Example of the Application of the Least Cost Model to Analysis ofSOz Damage to
          Painted Wood Residential Structures
Structure Use Life. 75 yrs.
Maintenance Strategy Repaint on first sign of damage.
S02 Level
(/jg/m3)
0
20
40
60
SO
WO
Number of Maintenance Net Present Value Maintenance Cost
Actions Over Use Life per Structure at Discount Rate of:

8
9
9
10
11
11
5%
3800
4100
4400
4700
5000
5200
10%
1600
1800
1900
2100
2200
2300
15%
900
1000
1100
1200
1300
1400
Structure Use Life: 75 yrs.
Maintenance Strategy: Repaint only after wood rot appears.
S02 Level
ffjg/m3)
0
20
40
60
80
100
Number of Maintenance
Actions Over Use Life
5
5
5
5
5
6
Net Present Value
per Structure at
5%
5000
5300
5500
5700
5900
6200
Maintenance Cost
Discount Rate of:
10% 15%
1800 800
1900 900
2000 900
2100 1000
2200 1100
2300 1100
Structure Use Life: 75 yrs.
Maintenance Strategy: Replace structure on collapse.

        SOz Level        Number of Maintenance Net Present Value Maintenance Cost
           (ug/m3)          Actions Over Use Life   per Structure at Discount Rate of:
                                                  5%           10%        15%
0
20
40
60
80
100
1 7800
1 8000
1 8200
1 8300
1 8500
1 8600
1000
1100
1100
1200
1200
1200
200
200
200
200
200
200
presence of SO2 are fairly well defined.
Unfortunately neither the physical damage
function  for  oaint nor the physical
damage functions for soiling are adequate
to characterize damage.  Information
gathered in these areas will improve the
accuracy on reliability  of estimating the
potential benefits and costs due to
changes in  ambient  concentrations of
atmospheric pollutants.
Edward F. McCarthy, Alexander Ft. Stankunas, and John E. Yocum are with
  TRC—Environmental Consultants,  East Hartford, CT 06108; Douglas Rae is
  with Charles River Associates, Boston, MA 02116.
Fred H. Haynie is the EPA Project Officer (see below).
The complete report, entitled "Damage Cost Models for Pollution Effects on
  Material." (Order No. PB 84-140 342; Cost: $14.50. subject to change) will be
  available only from:
        National Technical Information Service
        5285 Port Royal Road
        Springfield, VA 22161
        Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
        Environmental Sciences Research Laboratory
        U.S. Environmental Protection Agency
        Research Triangle Park. NC 27711

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
                              1L
                                                                                             t, US. GOVERNMENT PRINTING OFFICE: 1984-759-102/871

-------