An environmental protection publication in the solid waste management
series (SW-148).  Mention of commercial products does not constitute
endorsement by the U.S. Government., Editing and technical content of
this report were accomplished by the Resource Recovery Division of the
Office of Solid Waste Management Programs.

Single copies of this publication are available from Solid Waste
Management Information Materials Distribution, U.S. Environmental
Protection Agency, Cincinnati, Ohio 45268.

 by  Frank A.  Smith*
                         Background and  Purpose

      With  the increasing interest in  resource recovery  from municipal
 solid waste,  many state, city,  and regional  governments are seriously
 reevaluating  their solid waste  disposal  options.   In  the  process of
 evaluating both  design  alternatives for  processing  plants and potential
 revenues from recovered materials, these agencies  are finding that
 Information on the quantity and composition  of solid waste collection
 1s more Important to  the economics of system design for resource recovery
 than  was the  case for traditional  landfill and incinerator options.

      As we perceive more Interest in  such information, we also see that
 a confusing variety of  solid waste quantity  estimates are in circulation,
 Including  significant differences  among  "official" OSWMP figures.  For
 example, the  widely used "old"  figure of 5.32 Ib/cap/day = 190 million
 tons  (in 1967),  which emanated  from the  preliminary results of the 1968
 National Survey,  is contradicted by the  "new" figure of 3.32 Ib/cap/day =
 125 million tons  (in  1971),  which  has recently been used in papers,
 speeches,  and congressional  testimony by OSWMP personnel.

      This  paper  has two purposes.  The first is to provide definition
 and context for  both  the "old"  and the "new" figures, in order to
 minimize confusion and  to point out that,  in this business, there 1s
 no absolutely "correct  estimate".  The second purpose is to emphasize
 the necessity for extreme caution .in  applying any_ National average
 estimate to regional  or local solid waste management or resource recovery
 planning and  design efforts.

               A Review  of Selected Solid Waste Estimates

     The 1968  National  Survey.  The 1968 National Survey of Community Solid
Waste Practices represented a considerable effort to gather data from
     *Dr. Smith is an economist with the Resource Recovery Division
(RRD), Office of Solid Waste Management Programs (OSWMP), U.S.
Environmental Protection Agency (EPA).

field sources on a nationwide basis.  Returns were ultimately received
from communities representing 87 percent of the U.S. population.  In
retrospect, the Survey's major strength was its breadth of coverage, and
its major weakness was felt to be that the communities did not have good
data to provide.  In particular, the vast majority of communities reported
estimated rather than measured (weighed) data on solid waste.  Coverage
of commercial and industrial waste was especially spotty.

     The final returns were never fully evaluated, and the.principal
published solid waste figures from the Survey were the preliminary
results presented by Anton 0. Muhich and Richard D. Vaughan at the 1968
October meeting of the American Public Works Association.1  This was
the origin of the 5.32 Ib/cap/day = 190 million tons of solid waste
figure which has been so widely used both within and outside EPA.
It should be noted that these figures supposedly include reported demo-
lition, construction, industrial, and other municipal solid waste, in
addition to household, commercial, and institutional refuse of all kinds.
Of the 5.32 Ib/cap/day, only 4.15 was estimated as the household and
commercial contribution to the nationwide average.

     Paul Britton, a former OSWMP statistician, .subseguently conducted a more
rigorous evaluation of the complete returns from the 1968 Survey.
Figures based on his best estimates from a selected sampling of the
returns are presented in Table 1.  For various reasons, it was difficult
to establish a single best estimate for waste categories other than
household, commercial, and institutional, but the ranges presented are
nevertheless of interest.

     The most serious criticism of the ^1968 Survey is that it represented
estimated rather than measured (weighed) data.  Therefore, it must
be regarded with great skepticism.  Resource Planning Associates, Inc.,
 RPA), in-work performed for the National Center for Resource Recovery,
 NCRR),    has pointed out that the cities which reported measured
.weighed) data showed consistently lower per capita values for virtually
all waste categories than did the other respondents.  Although we con-
sider the sample size of communities reporting measured data to be too
small to consider their per capita values valid for National estimating
purposes (as was done by RPA), we do consider this as prima-facie
evidence of a tendency of the survey to overestimate per capita waste.
Thus, while accepting the RPA/NCRR qualitative conclusion that the 1968
Survey data overestimated waste to some degree, we still question the
quantitative extent of the difference in household municipal waste
implied by NCRR's published figure of 115 million tons for the municipal
solid waste stream in 1972 (exclusive of industrial and demolition waste).

     A second kind of evidence supporting a thesis that the 1968 Survey
overestimated the household-commercial-institutional component of
municipal waste was first propounded by Darnay and Franklin in 1970.4
This was in reference to the apparent contradiction between the reasonably

                                 Table 1
 Total 1967 Collected
    (million tons)
Average Per Person

Commercial and Institutional


Other Municipal

Demolition and Construction


Low       High
Estimate  Estimate

    6       22

    5       19

   34      105



Low        High
Estimate   Estimate




     *Britton, P.  Environmental Protection Agency analyses of final
questionnaire returns received by the 1968 National Survey of Community
Solid Waste Practices.  Unpublished data/ May 1970.  Average per person
(pounds/day) values estimated by Britton were converted to total 1967
annual tonnages on the basis of estimated 1967 U.S. population of
198.7 million.
     +Includes street and alley cleanings, tree and landscaping, park and
beach, and catch basin solids.  Excludes sewage sludge.

"hard" estimates of the National quantity of paper and paperboard waste
(based on well-known, annually published data on production,  consump-
tion, and recycling) and the measured percentage of paper material
typically found in composition analyses.  Thus, If. 145 million tons of
residential and commercial solid waste were collected in 1967, and if.
35 to 50 percent of this waste were paper (as consistently shown in
numerous sample composition analyses), it would follow that there are
between 50 and 73 million tons of wastepaper.  However, since the
independently estimated quantity of wastepaper for 1967 was approximately
34 million tons (oven dry) and since only part of the difference could
possibly be accounted for as moisture content in the wastepaper fraction,
the conclusion is that either the 145 million tons is too high, or that
35 to 50 percent paper in the waste stream is too high.  Many would
contend that the "softest" number should yield in cases of contradiction.
In this case, the 145-million-ton (4.01 Ib/cap/day) figure is the
softest of the three pieces of data.  However, the question of how much
moisture there actually is in the paper fraction of collected waste,
plus the inherent ambiguities of most of the percentage composition
studies, combine to preclude precise estimates and comparisons based on
total paper.

     Three further conclusions are suggested relative to  the  1968 Survey. First,
the 1968 Survey results have not been fully appreciated either by OSWMP
personnel or by the public at large, because these results were  never
analyzed sufficiently and because the best analyses have  not  yet been
published.  Second, it is evident that the published'results  (i.e. 190
million tons and 5.32 Ib/capita (1967) figures} have often been  wrongly
interpreted as being limited to conventional household and commercial
types of refuse, rather than including the fuller range of construction,
demolition, other municipal refuse, sewage sludge, and industrial  process
waste.* Third, there are two kinds of evidence to suggest that  the  1968
Survey  returns and the nationwide estimates which were based  on  them
tended  to overestimate collected municipal solid waste, although neither
type  of evidence can provide definitive proof regarding the extent of

      1971  Private Sector  Collection Survey.  As part of a study  of the
private sector refuse collection industry, Applied Management Sciences,
Inc.,  (AMS)  under contract to the National Solid Waste Management
Association  on a grant from the EPA, developed National estimates of
      *For example,  Darnay  and  Franklin argued the  "wastepaper paradox"
 in the context of 190  million  tons of "municipal"  waste, rather than in
 terms of a more relevant lower number more consistent with the types of
 household-commercial waste which  provide the packer-truck samples for
 virtually all  composition  studies.

U.S. per capita waste generation  (.collection) for 1970.5  Using large
samples from private waste  haulers, scaled up to National level
totals on the basis of  the  estimated share of total customers covered by
the private sector survey,  AMS was able to estimate residential, commercial.
and industrial solid waste.  The  AMS National estimates, modified to conform
with our definition of  residential and converted to 365/day/year basis, are
shown in Table 2.

      Methodologically,  the AMS Private Sector study estimates for National
waste levels resemble those of the 1968 National Survey in that both
involve scaling-up of data  from respondents on the basis of assumed
population coverage.  To a  large  extent, the AMS results also involve the
calculation of weight data, based on counted or estimated volume collec-
tion data reported by respondents.

      The AMS estimate  for  the "commercial" source category for 1971
(58 million tons or 1.57 Ib/cap/day) is remarkably close to Paul Brttton's
estimate for 1967 (57 million tons or 1.57 Ib/cap/day).  There is also
considerable consistency between  the two studies with respect to the
other non-residential categories, inasmuch as the AMS figures fall within
the ranges of values estimated by Britton.  However, the AMS adjusted
residential collection  estimate (128 million tons or 3.46 Ib/cap/day) is
substantially higher, than that of the 1968 National Survey (88 million tons
or 2.43 Ib/cap/day).

      Residential Collection Studies by the Systems Management Division
of OSWMP.  There is general agreement that household (residential) waste
constitutes the largest  source category of municipal solid waste.  Two
new sources of complementary data on the quantity of residential collection
have recently become available, as a result of empirical work on solid
waste collection systems supported by the Systems Management Division (SMD).
One set of data, covering a number of residential collection routes in
each of eleven city or  county jurisdictions, was developed by ACT Systems,
Inc., for a twelve-month period during 1972-73 under the Data Acquisition
and Analysis Program (DAAP) of SMD.  The other SMD-supported study was
conducted by Applied Management Sciences, Inc., and involved analysis of
over 20 communities during  the 1971-72 period.

      To a significant degree, both these studies avoided most if not all
of the shortcomings of  the municipal solid waste data previously available.
For example, both sets of data relate primarily, if not exclusively, to
residential sources, thus avoiding much ambiguity in interpreting which
sourcefs) the reported data actually apply to.  In addition, both sets of
      (l) relate to similar recent time periods;
      (2) are based on a calendar year's waste generation cycle;
      (3) cover a wide diversity  of city sizes and geographic locations;

                                 Table 2
                            MANAGEMENT SURVEY
Sources                             Total 1971 U. S.     Average Per Person
                                    Waste Collection*
                                      (million tons)        (Pounds/Day)

Residential                               128                 3.46

Commercial*                                 58                 1.57
     Subtotal                              187                 5.03

Demolition and Other                        12                 0.32

Industrial                                  66                 1.73
     Grand Total                           265                 7.08
     *Assuming a 1970 U.S. population of 203.2 million.
     +The original AMS values presented in Table 3.3, p. 3.4 of vol. 2,
Analysis of Data, were in terms of waste per average collection day.
Assuming 5.5 collection days per seven day week, the figures presented
here were calculated by multiplying the AMS figures presented here were
calculated by multiplying the AMS figures by 5.5 = 0.786 to convert to
calendar days.                                7
     ^The AMS study included apartment buildings of five or more units
in the "commercial" category.  The figure presented here regroups this
waste into the residential category on the basis of data in Table 3.2,
p. 3.4 of vol. 2, Analysis of Data.
     ^Assumed also to include "institution" customers.

       (4)  reflect weighed,  as opposed to estimated,  collection quantities
           (on  either a city-wide or collection route basis).

       Previously unpublished integrated results from the two  sets  of
 Individual  community studies are presented in Table  3.   They  are to be
 regarded as both preliminary and subject to possible revisions, following
 further verification in individual  cases.

       Although we might wish to have ten times as  many  case study  cities
 in  order to increase our sample sizes, the data from these two studies  are
 the best available on-^total  per capita residential collection.  Thus, it
 is  all the more unfortunate not to  have material composition  analyses to
 complement them.

       Although the results  from these two separate sets of residential
 collection studies are strikingly similar with respect  to both average  and
 median values, we should be extremely cautious about leaping  to nationwide
 conclusions on the basis of these relatively small samples.

       RRD  Material Flows Estimates  for 1971.   Because of a recognized need
 for more comprehensive and  integrated estimates of the  material and product-
 source composition of municipal solid waste (especially in terms of absolute
 nationwide magnitudes as opposed to individual local  area percentages), an
 in-house effort was undertaken in 1973 by the Resource  Recovery Division
 to  develop quantitative estimates for major materials comprising the "post-
 consumer"  municipal solid waste streams.    At the  same  time,  a cross-
 classification was developed between materials and consumer product
 categories. The principal  objective was  to estimate the quantities of
 key elements in the waste stream relevant to resource recovery and source
 reduction.   The focus was restricted to the types  of solid waste generated
 by  the household and commercial-institutional  sectors (the latter  includ-
 ing office-type activities  of governments  and manufacturing enterprises).

       Thus, our scope encompassed approximately the  same household-
 commercial-institutional categories discussed in relation to  the 1968
 National Survey.  The RRD study excluded  all  reference  to the following
 categories: (1) industrial  processing waste;  (2)  construction or
 demolotion residuals; (3) street sweeping  (except  for assumed packaging
 material litter); (4) heavy or bulky tree and landscape waste other than
 yard waste accepted in ordinary collection; (5) sewage  sludge.  Also
 excluded were  any residuals  relating to automobiles,  auto-wrecking and
       *This  is  not invariably clear with  respect to  the AMS  data  sources,
which  require some further confirmation on  this  issue  as well  as  on  the
question  of  whether bulky waste is  included.   Bulkies  are  excluded for
the most  part 1n the ACT System data.

       +General  results, definitions, and other issues are presented in
ch  1  of the Second Report to Congress: Resource Recovery and Source
Reduction.  A paper presenting detailed methodology is available:
Smith, F.L.  A solid waste estimation procedure:  material  flows approach.
Environmental Protection Publication SW-147.  Washington, U.S. Environmental
Protection Agency, 1975. 56 p.


                                 Table 3
                          STUDIES  (LB/CAP/DAY)
                                    ACT Systems,       APPLIED MANAGEMENT
                                    Inc. (DAAP)          SCIENCES, INC.

                                    11 Communities,    17 Communities,
                                      1972-1973*         1971-1972*
Unweighted arithmetic average
  of individual community data            2.38               2.35

Median                                    2.43               2.40

Range of values                     1.72-3.43         1.10-3.40
     *Based on computer print-out data supplied by ACT Systems, Inc.,
February 1974.  For the most part, data are known to exclude bulky waste.
     +Data extracted from individual city reports by Martha Madison and
Cynthia McLaren, SMD, OSWMP.  Data for some cities may include bulky waste
and/or exclude significant portions of yard' waste.

and scrapping (tires, however, were  included  in the rubber category).  The
types of materials and products are  found  in  Table 4 on the following

      With respect to methodology, all  product-type waste (except food) was
estimated independently, on the basis of available published data for
specific material or-product  flows,  including production, international
trade, and marketing data, together  with estimates of average product life-
times and current recycling rates.   The non-food post-consumer product
waste (including "bulky" waste, such as major appliances, tires, and home
furnishings) estimated for 1971 is 77.1 million tons on an "as generated"
moisture content basis,* with bulky  waste  comprising about 7 million tons.
At this point,  1t is worth emphasizing  that our material/product estimates
originate on a  nationwide basis and, therefore, require no scaling.

      The remaining three categories of food, yard, and miscellaneous
inorganic waste were estimated indirectly  on  the basis of their respective
average percentages in typical municipal waste composition analyses as
calculated by W. R. Niessen and S. H. Chansky in their 1970 National
Incinerator Conference paper.+  Thus, food, yard, and miscellaneous
inorganic fractions (48 million tons) were estimated from the Niessen-
Chansky ratios  (Table 4).

      This briefly describes  the definition and origin of the 125-million-
ton figure, which, for a 1971 U.S. population of 207 million, implies a
per-capita figure of 3.31 Ib/day.  An additional step was then taken to
distribute the  material quantities into household and non-household source
components.  This was done for each  material, individually, by product
categories, and then aggregated into sector totals, by material (Table 5).
With the important exceptions of the paper product grouping and certain
of the non-paper container categories for  which we drew primarily on
Midwest Research Institute estimates of household versus non-household
waste generation rates, much  of this exercise was conducted on the basis
of our own intuitive judgement and previous experience rather than on
the basis of observation.  Nevertheless, it was conducted in as objective
a manner as possible, and the product-by-product approach guaranteed that
the resulting residential vs. non-residential totals would be completely
Independent in  derivation from other estimates based on waste collection
      *i e  , paper products were estimated on a 7 percent moisture "air
drv" basis''glass and metals on a zero-moisture basis; and the remaining
materials at 3 to 15 percent moisture.  The total 125-mi 11 ion-ton waste
flow is estimated to contain about 26 percent moisture overall.
      *The  actual calculations weee based on our estimate of 70 million
tons of non-bulky materials, since these were felt to be more closely
comparable  in composition to the typical packer-truck loadings, upon
which most  percentage material composition sampling has been based.

                                                   TABLE 4

Kinds of Materials
Other Nonferrous
Rubber & Leather
Non-Food Product Total s
Food Waste
Yard Waste
Misc. Inorganics

"As Generated
u»T3  I- O!
a to  c
0. C C T-
IO *•»- M-*O Ol
.^ xx ig -i_m<\x

                                                  TABLE  5

                              GENERATION,  BY WASTE MATERIAL CATEGORY,  1971*

Non-bulky Product Waste:
Rubber & Leather
Subtotal: Non-food
Food Waste
Subtotal : Non-bulky
Product Waste
Yard Waste
Misc. Inorganics
Total Non-bulky Mat'l
Add: Bulky Waste
Total Residential
Weight Basis
106 Tons
0.169 '

Weight Basis
, 106 Tons
Weight Basis
10° Tons . Lb/ nap/day . Percent

     *Smith, F. A., and F. L. Smith, Resource Recovery Division, Office of Solid Waste Management Programs,
Feb. 1974.

      Before comparing the results achieved in this manner with those of
the studies previously discussed, a further word is needed concerning
the accuracy of such RRD estimates.

      We previously argued that the 1968 Survey overestimates collected
waste.  Similarly, there should be nothing sacrosanct about the RRD
material-flows estimate of 125 million tons.  Basically, there are at
least three errors that could have occurred in the rather'involved
estimating method.  Briefly, they are:  (1) incorrect basic estimates for
individual non-food product materials; (2) failure to include one or more
significant product-related materials entirely; (3) application of incorrect
percentage factors in estimating food, yard, and miscellaneous inorganic

      This is not the place to argue the details of any of the three
possibilities for error.  It suffices to say that we have no way of knowing
if our present estimates are too high or too low with respect to errors
(1) and (3).  That is, while our non-food product flow estimate (77 million
tons) could easily be incorrect by 10 percent or more in the aggregate,
the error could be in either direction.  Resource Planning Associates
independently performed much the same procedure as we did for estimating
non-food product waste and estimated approximately 10 million tons less.7
Because we have confidence in the two largest non-food material/product
estimates—those for paper and glass which together comprise two-thirds
the weight of our total non-food product estimates—we have more confidence
in the non-food product sum of 77 million tons (2.04 Ib/cap/day) than in
the other categories.  As previously noted, the estimating factors we used
to calculate food and other non-product fractions are based on the work by
Niessen and Chansky in synthesizing previous percentage composition studies.
We feel that the ultimate data base here is indeed quite ambiguous and
weak, but still, we should not expect much more than a 25 percent error in
either direction for the sum of these three indirectly estimated categories.*

      If any major tonnage materials or products have been omitted, these
would cause us to underestimate total waste.  We feel that we are more
likely to have overlooked a number of small items than any single large
Item, and we are somewhat less confident regarding the non-paper fraction
of commercial-institutional categories than in the household sector
      *In a 1972 extension of the 1970 work, Niessen and Alsobrook7 revised
the as-discarded weight-percentage contribution of food, yard, and mis-
cellaneous inorganics downward somewhat from 40.7 percent to 37.3 percent
of (non-bulky) total watte, based on additional composition studies.  Use
of these later factors would reduce our total waste estimate by about
6.5 million tons.

      It should be noted in this regard that any overestimate or under-
estimate in the non-food product categories will be compounded by a factor
of about 1.67 due to our percentage method of estimating food, yard, and
miscellaneous fractions.*
                  Comparative Results and Conclusions
         for The Residential/Commercial-Instituti'onal Sectors
      We now have three  independent estimates for the total quantity of
residential and commercial solid waste flows:  the 1968 National Survey
(1967 data), as evaluated by Paul Britton; the 1970 figures from the AMS
Private Sector study;  the RRD materials flows estimates for 1971.  The
three sets of data are summarized in Table 6, with both the 1968 National
Survey and the AMS Private Sector data "normalized" to 1971 by assuming
growth in waste collection over the relevant periods equivalent to growth
in  total National real expenditures on non-durable consumption goods.

      The following  comparative results for these three sets of 1971
estimates are apparent from Table 6:

       (1) total waste  varies from a low of 125 (RRD) to a  high of 191

       (2) the 1968 National Survey and RRD estimates are in reasonably
          close agreement—within about an 8 percent difference—for
          residential  waste.   But RRD shows only 55 percent of the
          National Survey's 64 million tons of commercial-institutional,
          which accounts for most of  the  overall difference of 36 million

       (3) also, the  National Survey is in very close agreement with  the
          Private  Sector Study on Commercial-Institutional waste (within
          7  percent),  but there  is  a  34 million ton difference  in the
          residential  sector estimates;

       (4) the  RRD  figures are  more  than  30  percent  lower for  residential
          and more than  40 percent  lower  for commercial-Institutional  than
          the  Private Sector  Survey estimates, with a  total difference of
          about 66 million tons.
       *It should be possible, in principle, although we have not yet done
 so  to develop a better, more direct empirical estimate for food waste
 ,j?ina the material flow approach as a check on waste-collection samples.
 Improvement in our estimate of this important category would significantly
  Crease our confidence in the overall total estimates.  Improvement
 estimates for yard waste must rely on further direct field sampling.

                                   Table 6
                            SOLID WASTE FOR 1971
                              1968 National Survey* AMS Private    RRD Materials
                                                    Sector Survey*     Flows^

Residential 97.5
Commercial-Institutional 63.8
Totals 161.4

Residential 2.58
Commercial-Institutional 1.69
Totals 4.27




     *From Table,!, adjusted upwards by 11 percent, to account for growth
equivalent to the change in real U.S. personal consumption expenditures on
non-durable goods between 1967 and 1971.
     +From Table 2, adjusted upward by 2.5 percent growth factor for 1971
over 1970.
     "'"From Table 5.

       Residential.   Of  the  three estimates of residential solid waste
summarized  in Table  6,  the  two  lower values  (1968 National Survey and RRD
oaonr}?/  Flows) were seen to  be ^  reasonably close agreement at 2.58 and
2.39  Ib/cap/day.  Lower values  are  supported by the results of both
residential  collection  studies  (Table  3).  Both the average and median
values of the 17-community  AMS  data and  the  11-community ACT Systems data
are grouped  within a few percentages of  one  another, with averages at
2.35  and  2.38 and medians at  2.40 and  2.43 Ib/cap/day, respectively.

       Both  the AMS and  ACT  Systems  data  are  presumed to relate to non-bulky
residential  collection,  either  wholly  or primarily.  On this basis, they
should be increased  by  about  0.15 Ib/cap/day to correspond with the RRD
estimate, which explicitly  includes household bulkly items, and the other
two estimates, which implictly  include them.  On the other hand, the two
residential  collection  studies  represent data for a somewhat later time
period.   Adjustment  for this  factor would reduce these values somewhat
(2 to 4 percent).  The  net  effect should be a slight upward adjustment of
less  than 0.1 Ib/day for purposes of comparison against 1971 residential
figures,  which include  bulky  product waste items.

       These  four sets of data on residential waste—the 1968 National Survey,
the RRD Material Flows  Analysis,  and   the two (small sample) residential
collection  studies—when adjusted to a common 1971 basis, thus estimate
the residential waste fraction  at between 2.4 and 2.6 Ib/capita/day, includ-
ing moisture.

       There  is also  further evidence that the "true" National average lies
at a  point close to  these values.   The best estimate that we have for
wastepaper generated in  the residentia-1  sector is a National average for
1971  of 0.57 Ib/capita/day  (after recycling), on an oven dry basis.*  Our
best  data on the percentage of  paper in  residential collection come from
a three-city study conducted  by ACT Systems.9In these three cases, the
annual  average percentage varied only  between 29.5 and 31.0.  We do not
have  good data on the percentage of moisture in collected wastepaper, which
is an important variable in any attempt  to reconcile material flow
estimates with waste collection values.  However, if residential waste-
paper collection averages somewhere between 20 and~3~0 percent moisture
(on a wet weight percentage basis), then the wet weight of residential
wastepaper generation lies  somewhere between 0.71 and 0.82 Ib/capita/day.
If paper  is  in fact  about 30  percent of  residential waste on average, then
a value of wastepaper between 0.71  and 0.82 Ib/capita/day implies a
National average residential  waste  between 2.38 and 2.7 for 1971 ."*"
      *See Table 5.
      +For the lower moisture percentage C20 percent of wet weight of
oaoerV  0 57 lb- 0.3 = 2.375 Ib/cap/day.  For the higher moisture level
[30 percent):  0.57 Ibf o.7~ .3 = 2.714 Ib/cap/day.  If paper is more
than 30 percent of residential waste, then total waste projected from a
given wastepaper generation rate will, of course, be lower than the values
estimated here.

      On this basis, values of total residential waste higher than
2.7 or 2.8 Ib/capita/day would require that the paper being collected is
extremely wet—more than 30 percent or so moisture—or that paper is a
lower fraction of residential waste than we think it to be—i.e., lower
than 29 percent.  A National average value of 3.5 Ib/capita/day is
inconsistent with existing evidence on the quantity of household waste-
paper generation, the percentage of paper typically observed in residential
collection, and the types of motstyre levels typically recorded in
residential waste.  Additionally, a value of 3.5 exceeds the highest
values measured in either the ACT System or the AMS residential collection

      Our conclusion on residential waste collection is that the National
average value for 1971 on a wet weight basis very likely lies within the
range of 2.3 to 2.7 Ib/capita per day (87 to 102 million tons).  Since
we do not know as much as we would like to about either the average
moisture content of collected waste or the average moisture content of
its individual constituents, more precise reconciliation than this seems
impossible at present, given the existing kinds and sources of data and

      Commercial-Institutional.  We have seen considerable divergence
between the RRD materials-flow estimate of 35 million tons (0.92 Ib/capita/
day) of commercial^institutional waste generation and the two survey
estimates whose time-adjusted values fall between 60 to 64 million tons
(1.6 - 1.7 Ib/cap/day).  We lack independent field study'estimates based
on measured (weighed) collection data to serve as solid evidence on this
subject.  Nonetheless, we cannot justify estimated collection as high as
60 million tons, on the basis of our material-flow data.

      Little work has been done  to characterize the material composition
of  commercial-type waste, but the consensus appears to be that commercial
and  institutional sources have a very high percentage of paper—higher
than 50 percent and perhaps as high as 75 percent on the average.  Although
there is no comprehensive empirical basis for determining precisely what
the  percentages of paper in the  commercial waste stream is, our material-
flow estimates  indicate that it  is unlikely that any more than 15 million
tons of non-recycled wastepaper(oven-dry weight) was generated by all
non-residential sources in 1971.  At 25 percent moisture when collected,
the  paper  fraction would then weigh about 20 million tons.  If paper were
only 50 percent of the commercial sector waste, then 40 million tons of
total waste would be  projected for that sector  in 1971.

      We think  it unlikely that  this sector's wastepaper would have an
average moisture content higher  than 25 percent, or that paper would
constitute less than  50 percent  of the wet weight of collection from this
sector.  Therefore, on the basis of this type of evidence (inferential
and  circumstancial as  it is) it  is difficult to understand how 1971
commercial sector waste collection  (as we define it) could be much higher *
than 40 million tons,  or about 1.06 Ib/capita/day as a National average.

      Aside from possibilities of overestimation on the part of question-
naire respondents or possible errors involved both in converting reported
volume rates into tonnage terms and sample population data into total
population data, the most logical explanation for high collection-based
estimates vis-k-vis low material flow-based generation estimates for the
commercial sector is differing definitions of the sector in question.
Thus, questionnaire respondents or interviewees may, in some instances,
have interpreted the term "commercial" to include a broader range of
establishments overlapping into some industrial processing and manufac-
turing assembly categories.  In addition, collection routes that were
predominantly commercial .may, as a matter of convenience, have been reported
as entirely commercial.  Finally, there may have been some reporting of
materials collected which were subsequently salvaged for recycling rather
than landfilled or incinerated.  All this is, of course, speculative.  The
simple fact is that we do not presently have a satisfactory or "scientific"
explanation for the 20 to 25-mi11 ion-ton-difference in estimates for the
commercial sector.

1.    Black, R. J., A'. J. Muhich, A. 0. Klee, H. L. Hickman, Jr., and
        R. D. Vaughan.  The national solid wastes survey; an interim
        report.  [Cincinnati], U.S. Department of Health, Education,
        and Welfare, 1968.  53 p.

2.    Resource Planning Associates, Inc.  Potential economic value of
        the municipal solid waste stream.  Washington, National Center
        for Resource Recovery, Inc., Sept. 1972.  19 p.

3.    Municipal solid waste; its volume, composition and value.
        NCRR [National Center for Resource Recovery] Bulletin,
        3t?]~: 4-13, Spring 1973.

4.    Darnay, A., and W. E. Franklin.  Salvage markets for materials in
        solid wastes.  Environmental Protection Publication SW-29c.
        Washington, U.S. Government Printing Office, 1972.  187 p.

5.    Applied Management Sciences, Inc.  The private sector in solid
        waste management; a profile of its resources and contribution
        to collection and disposal,  v.2.  Analysis of data.  Environmental
        Protection Publication SW-51d.l.  [Washington], U.S. Environmental
        Protection Agency.  1973.

6.    Niessen, W. R., and S. H. Chansky.  The nature of refuse,   ln_
        Proceedings; 1970 National Incinerator Conference, Cincinnati,
        May 17-20, 1970.  New York, American'Society of Mechanical
        Engineers,  p. 1-24.

7.    Niessen, W. R., and A. F. Alsobrook.  Municipal and industrial
        refuse; compositions and rates.  In_ Proceedings; 1972 National
        Incinerator Conference, New York, June 4-7, 1972.  New York,
        American Society of Mechanical Engineers,  p. 319-337.

8.    A summary of the characterization of residential solid waste for
        the period of April 1971 through June 1972.  Winter Park, Fla.,
        ACT Systems, Inc., Aug. 15, 1972.  19 p.
      i>U.S. GOVERNMENT PRINTING OFFICE: 1977- 241-037:27