,- rť ,-,
530SW138C
Energy Potential
from
Construction and Demolition Wood Wastes
by
JACA Corp.
Environmental Consultants § Engineers
506 Bethlehem Pike
Fort Washington, Pennsylvania 19034
Contract No. 68-01-3560
John W. Thompson, Project Officer
Prepared for
Office of Solid Waste Management
U.S. Environmental Protection Agency
Washington, D.C.
February 1977
-------
PREFACE
Energy shortages and the need to conserve natural resources have led
the United States to examine its waste streams as sources of potential supplies
of energy and reconstituted products. While solid waste energy potentials in
the private sector have been examined in some detail, construction and demoli-
tion wastes have not heretofore received comparable study attention. Therefore,
the first step in analyzing a reuse program is the development of the size and
nature of the particular waste stream.
The objective of this report was to determine the nationwide volume of
construction and demolition wastes, and, in particular, the combustible fraction
of wastes from these activities. Ancillary outputs of the study are annual
generation rates by location, information on time variation in waste flows, and
the distribution of wood fractions.
Appreciation is expressed to the National Association of Demolition Con-
tractors, their president Mr. Ron Dokell, Mr. William Baker, Executive Secretary
and members of the NADC Energy Recycling Committee for providing valuable assis-
tance in scheduling fieldwork and for the administration of a response card
program regarding volumes of waste generated per demolished building. We are
also grateful to Mr. Mike Sabia of Domino Salvage who assisted us in the training
phase of the program, and Mr. William Geppert of Geppert Brothers who helped in
some of the original planning aspects of the work.
The twelve month study was conducted by JACA Corp. of Fort Washington, PA
under contract number 68-01-3560 with the Office of Solid Waste Management Pro-
grams of the Environmental Protection Agency in 1976. The EPA project manager
was John W. Thompson. The principal investigator was James A. Commins.
Researchers were H. Baist, G. Gindlesperger, G, Lester, E. Mathis, and W. Ott.
-------
Table of Contents
Page
SUMMARY 1
RECOMMENDATIONS 2
BACKGROUND 3
NATIONAL AND AREA ESTIMATES 4
ECONOMICS OF WOOD RECOVERY 16
Wood Fraction as a Fuel or Fuel Supplement. . 17
Wood Fraction as Pulp Fibers 17
Wood Fraction as a Building Material
Substitute 18
PILOT PLANT OPERATION 19
METHODOLOGY § FIELD DATA 20
Essentials of the Methodology 21
Training 22
Statistical Methodology 26
Stratifying the Sample 32
Collecting Field Data 39
Density Determination 41
Volume Determination by Response Cards .... 43
Wood Waste Generated by Construction Activity. 47
Estimating the Total Number of Buildings De-
molished Per Year and Constructed Per Year
in the United States 47
Estimating the Energy Potential from Demo-
lition and Construction Waste on National
and Local Levels 49
DISCUSSION OF ACCURACY OF RESULTS 52
Demolition Waste 52
Construction Waste 55
REFERENCES 57
-------
List of Tables
Table Page
1 ANNUAL TONS AND BTU POTENTIAL OF WOOD WASTE
FROM CONSTRUCTION AND DEMOLITION WASTES OF
10 CITIES IN 1976 5
2 PROJECTED CONSTRUCTION RATES FOR RESIDENTIAL
UNITS 11
3 NOMINAL HEATING VALUE OF VARIOUS WASTES AND
FUELS 15
4 COMPARISON OF ANNUAL BTU POTENTIAL 15
5 SUMMARY OF TRAINING RESULTS 28
6 DEMOLITION AND CONSTRUCTION FIELD DATA
STRATIFICATION 36
7 CALCULATION OF PERCENT COMBUSTIBLE DEMOLITION
WASTE, PHILADELPHIA, PA 37
8 WEIGHTED PERCENT OF WOOD IN DEMOLITION AND
CONSTRUCTION WASTE 42
9 CITIES REPRESENTED IN VOLUNTEER RESPONSE CARD
PROGRAM 46
10 ANNUAL BTU POTENTIAL OF WOOD WASTE FROM DEMOLI-
TION AND CONSTRUCTION WASTE FOR TEN CITIES . . 51
11
-------
List of Figures
Figure Page
1 RECENT HOUSING UNIT DEMOLITION RATES IN
MAJOR U.S. CITIES 7
2 CONSTRUCTION AND DEMOLITION ACTIVITY IN
PHILADELPHIA, 1975 8
3 CONSTRUCTION AND DEMOLITION ACTIVITY IN
PORTLAND, 1971-1975 9
4 U.S. BUILDING CONSTRUCTION TRENDS 10
5 U.S. WOOD CONSUMPTION AND SHIPMENTS TO BUILDING
AND CONSTRUCTION 12
6 PERCENT WOOD IN CONSTRUCTION MATERIAL
CBY WEIGHT), 1951-71 13
7 FREQUENCY DISTRIBUTION OF PERCENTAGE
COMBUSTIBLE DATA 14
8 DIAGRAM OF TRAINING AREA - DOMINO SALVAGE -
PLYMOUTH MEETING, PA 23
9 TYPICAL 40 YD.3 LOAD OF DEMOLITION WASTE AT
LANDFILL 24
10 AN UNSEPARATED LOAD OF DEMOLITION WASTE WITH A
HIGH PERCENTAGE OF WOOD 24
11 AMOUNT OF WOOD HAND-SEPARATED FROM DEMOLITION
WASTE IN 1.5 HOURS DURING THE TRAINING PERIOD. 25
12 FREQUENCY DISTRIBUTION OF ACTUAL COMBUSTIBLE
PERCENTAGE -LOADS 1 to 32- 27
13 AVERAGE ABSOLUTE ERROR PER TRUCKLOAD 29
14 GEOGRAPHICAL DISTRIBUTION OF CITIES INCLUDED
IN THE SURVEY 34
15 NUMBER OF RESPONSE CARDS COMPLETED BY
RESPONDENTS 45
111
-------
SUMMARY
Based on a 1976 survey of 10 cities, the annual volume of wood wastes
from construction and demolition wastes is 2.37 million tons. Projected to
a national basis by statistical means, the annual tonnage of wood waste from
demolition and construction waste is estimated at 21.9 million tons annually.
Demolition wood wastes accounts for 19.3 million tons and construction wood
wastes 2.6 million tons of this total.
If the energy potential of these wood fractions could be recovered it
would have an approximate heating value of 3.75 x 1C)14 BTUs. This heating
value is equal to about 16 percent of the forest and industrial waste stream
or 2 percent of the total coal supply.
Several significant aspects about the waste stream from the demolition
and construction industry resulted from the study:
(1) The average density of construction and demolition debris at dis-
posal sites is 25 pounds per cubic foot.
(2) The average load of demolition debris contains about 39 percent
wood and construction wastes 48 percent wood. However, these averages are
deceiving because loads tend to run either 10 percent wood or 80 percent wood
content.
(3) The average duration of a demolition job is 3.87 days for residen-
tial, 14.7 for industrial, and 9.56 for commercial buildings.
(4) The average volume of a demolition job is 450 cubic yards for resi-
dential, 3,856 for industrial and 2,022 for commercial.
(5) The percentage of wood by weight with respect to other construction
materials has been steadily decreasing over the last 25 years to a low of 9.8
percent in 1971. Estimates of wood waste in the construction industry averaged
7.4 percent of deliveries.
(6) High risk and entrance costs due to separation problems, sporadic
supply and undeveloped markets stymie the reuse of demolition and construction
wood wastes. However, there are two small firms in California marketing pre-
screened construction and demolition wood waste to the pulp industry.
-1-
-------
RECOMMENDATIONS
(1) The energy and/or product potential of demolition and construction
wood waste is sufficiently large to merit further investigation on two criti-
cal points - separation techniques (and costs) and market potentials.
(2) Risk and rate-of-return characteristics however are such that it
is doubtful that private enterprise will undertake a program to reuse such
wastes at present energy costs. It is recommended that the federal government
consider the feasibility of lessening the risk and improving the rate-of-return
expectations so that private parties will enter this market.
C3) The seldom used concept of separate dump sites for demolition and
construction waste should be examined for technical and economic viability.
-2-
-------
BACKGROUND
Wood waste from primary and secondary manufacturing and from logging
operations have been estimated at 16 million tons and 118 million tons respec-
tively.^ This estimated waste is more than two times current U. S. total pulp-
wood fiber needs. Based on this study, the amount of wood waste available from
demolition and construction waste are of the same order of magnitude. The total
waste from wood processing plants and logging forest waste is 134 million tons
versus 21.9 million tons for demolition and construction waste.
Some of the process and forest wastes have been put to energy saving
uses. Applications have included direct use of these wastes as fuels or as
fuel supplements. Indirect energy savings uses have been for mulch, medium
density particle board, cattle bedding^ and grading for recreational areas^.
An industry has grown up around this reuse of wood wastes. There are more
than twenty manufacturers6 of equipment for the handling and/or disposal of
trees and wood wastes. Of these, however, seven are manufacturers of air
current destructors, whose primary purpose is to reduce the volume of waste
by burning without recovering the heat.
Plants that use large quantities of wood in fabrication of finished
products often use the wood as a fuel source or supplement to produce heat
or steam for local operations. There are a wide variety of furnaces especially
designed for burning wastewood fuel. The variety of furnace techniques is
quite broad, and includes fluidized beds, inclined grates, dutch oven,
spreader-stoker, fuel cell, hogged fuel and sawdust burners, sander-dust
burners, etc.
The greatest use of wood for fuel or a fuel supplement takes place at
these wood fabricating plants, where a constant supply of wastewood is avail-
able and an on-site fuel need exists. The situation here is analogous to
"home scrap" where there is a ready raw material supply and a local need.
While the woodwaste from demolition and construction waste estimated in
this study is of the same order of magnitude as that from municipalities and pro-
cessing plants, there are certain peculiarities in its generation and composition
which aggravate problems associated with its productive reuse or direct use as a
fuel or fuel supplement.
Demolition waste is generated when buildings are demolished because of fire,
unsafe conditions, or to make way for other buildings or alternate land use. The
business is served by more than two thousand demolition contractors, most of them
small operators. Demolition jobs are short lived. This study showed average job
spans of 3.27 for residential, 9.56 days for commercial buildings, and 14.7 days
for industrial buildings. Choice of a dump site is governed by optimizing the
combination of dumping fees and hauling expenses. Hauling is customarily done by
the demolition contractor often using special trucks to facilitate loading and
unloading bulky materials.
-3-
-------
The generation of construction waste is quite different. It is generated
from cutting material to size and from breakage and wear out of materials such
as concrete forms and scaffolding in the construction of buildings. A construc-
tion job is a much longer term project than a demolition job, ranging from a
month to more than a year. General contractors seldom haul their o\m waste.
Usually they contract with private haulers who utilize bins which are left at
the site and periodically removed. Often each sub-contractor is responsible for
removing his waste, and this gives rise to a multiplicity of storage and hauling
arrangements.
The amount and source of wood waste generated varies considerably during
the several phases of constructing the foundations, shell and interior.
The common thread in both types of waste is that they are highly hetero-
genous both as to composition and size. Neither contain putrescibles, but both
can contain brick, stone, metal, glass, wood, impregnated paper, mortar, cinder-
block, and other detritus. Size of the waste can vary from small wood fragments
to large chunks of reinforced concrete, timber, sheets of metal roofing, etc.
The waste is highly varied insofar as size, weight, density, hardness, and other
physical attributes.
If this waste is to be seriously viewed as a material resource or a fuel,
characteristics of its generation and present disposal must be better understood
in terms of amounts, predictability of supply, use technology and use economics.
This study focused on the first two aspects of the problem - amounts and predicta-
bility of supply.
NATIONAL AND AREA ESTIMATES
Construction and demolition activities were studied in ten cities; Phila-
delphia, Los Angeles, Chicago, Houston, Detroit, Miami, St. Louis, Atlanta,
Pittsburgh, and Minneapolis with the emphasis on determining the combustible
fractions present in these wastes at area landfill sites. Demolition and construc-
tion rates for each city as well as heating values for various wastes and fuels
were analyzed to determine the potential energy represented by these combustible
fractions.
Tonnages of construction and demolition wastes and the nominal energy
potential vary considerably by city and region of the nation. Of the ten
cities studied, Atlanta exhibited the lowest annual potential, .72 x 10 BTUs,
while Los Angeles, Chicago and Detroit were clustered at the high end; Houston,
Miami, Minneapolis, Pittsburgh, and Atlanta at the low end; and Philadelphia
and St. Louis near mid-range on Table 1.
-4-
-------
Table 1
ANNUAL TONS AND BTU POTENTIAL OF WOOD WASTE
FROM CONSTRUCTION AND DEMOLITION WASTES
OF 10 CITIES IN 1976
City
Philadelphia
Los Angeles
Chicago
Houston
Detroit
Miami
St. Louis
Atlanta
Pittsburgh
Minneapolis
Thousand Tons
190
670
400
110
490
67
230
42
100
67
BTU Potential
3.3 x 10
12
11.0 x 10
12
6.9 x 10
12
1.9 x 10
12
8.5 x 10
12
1.1 x 10
12
4.0 x 10
12
.72 x 10
12
1.8 x 10
12
1.2 x 10
12
Average
237
4.0 x 10
12
-5-
-------
Demolition rates vary from city to city, as can be seen on Figure 1.
Source cities also display large variations from year to year, viz. Detroit,
St. Louis, Cleveland, Dallas, and Kansas City. Construction and demolition
activity in Philadelphia and Portland respectively are shown on Figures 2
§ 3 as a function of month. Both indicate, as one would expect, lower demo-
lition activity than construction. Both graphs show that construction and
demolition activity can have swings for a year of 100 percent. This is not
apparent by a cursory comparison, since the level of construction is so
much greater than the level of demolition activity. Several trends in
building construction rates and construction materials are discernable:
Construction rates of non-residential units from 1957 through
1974 were less volatile than residential unit rates. Both
rates were , in 1974, about equal to those of 1957 (Figure
4).
Construction of residential units is expected to move to a
higher level through 1981 (Table 2).
Building construction is on the increase thus wood consump-
tion in the building and construction industry continues to
increase (Figure 5). Shipments of wood to U. S. building and
construction have increased about 30% from 1951 to 1971.
However, wood used per building as a percent by weight of
other materials has experienced a persistent decline from
13.5% in 1951 to 9.8% in 1971 (Figure 6).
Recent technical trends in the building industry have reduced
the amount of wood per structure. For example, modern
scaffolding uses more pipes and clamps than the older all
wood scaffolds. Also the wooden forms used in concrete work
are being replaced by plastic or metal forms. Moreover,
there is an increasing amount of metal, especially aluminum,
being used in metal frames, door jambs, and siding in commer-
cial and industrial buildings.
At the disposal sites, the average percent combustibles in demolition
waste was 39% and the average in construction waste 48%. The reason these
averages tended toward the midrange is due to the "U" type distribution of the
observed waste loads, where waste compositions were clustered at the high and
low ends of a distribution curve. A frequency plot of the distribution based
on 1001 field observations is shown in Figure 7.
The nominal heating value of various fuel sources are shown in Table 3.
The heating values, of course, cover a range but those nominal values shown
were chosen for comparison purposes. The heating value of wood is about mid-
way between fuel values from commercial, residential and industrial wastes,
and that of coal. Using this value, and the 21.9 million tons of combustible
demolition and construction waste this study estimated that the total energy .
potential which would be generated annually would be approximately 3.75 x 10
BTUs.
-6-
-------
8000 ..
Figure 1
RECENT HOUSING UNIT DEMOLITION RATES
IN MAJOR U.S. CITIES
7000
6000
5000 -.
4000
3000
2000
1000 ..
Minneapolis
Jacksonville -"
Dallas
Cincinnati
Philadelphia
Indianapolis
\ Kansas City
1971
1972
1973
1974
-7-
-------
f-l
bC
H
t,
CU
a.
2
>
ii
E-
<
01
O
Q
O
i i
E-
U
oi
2
O
CJ
c
o
O
CD
Q
o
o
Q
>
O
u
O
t/J
<
0)
3
X
rt
CD
tu
C
cd
O
o
CN
O
o
1 (
o
o
o
o
o
o\
o
o
00
o
o
r^
o
o
vO
O
o
LO
O
o
^r
o
o
K)
O
o
CM
1
O
O
i-H
Units
-------
! i "urc .:>
CONSTRUCTION AND DEMOLITION ACTIVITY IN PORTLAND, 1971-1975
800
700
600
500
to
C
400
Construction 1975
Construction Avg. 1971-1975
300
>00
IOC ..
Demolition Ave. 1971-1975
Demolition 1975
Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.
-------
/ 1
O
r-H
X
CN
vO
O
i-H
X
LO
vQ
O
i l
X
LO
1
O
r-H
X
LO
CM
i
O
r (
X
fl
1
LO
O
I I
X
LO
r~-
i
LO
O
rH
X
LO
\
\
LO
O
r-H
X
LO
CM
'VS'fl P31onj;isuo3 SITUQ
-10-
-------
Table 2
PROJECTED CONSTRUCTION RATES FOR
RESIDENTIAL UNITS
Projected Residential Units
1976 1,512,900
1977 1,770,100
1978 1,444,000
1979 1,650,000
1980 1,790,000
1981 1,870,000
Source: National Association of Homebuilders
-11-
-------
LL
O
in
2:
o
10
kO-
so-
40-
30-
^ 10 -
O
if*
Figure 5
U.S. WOOD CONSUMPTION AND
SHIPMENTS TO BUILDING AND CONSTRUCTION
BUlLDiMG AND
CON5TRUCTIOW
to
CO
ir\
CO
Q
tr-
4-
ir\
r-
YEAR
Source: See Reference 2
-12-
-------
Figure 6
r-N
+->
M
H
0
X
f!
t 1
rt
H
?H
O
P
ctf
2
C
0
H
O
3
t/i
G
O
U
O
-3
O
O
S:
\ j
O
0)
PERCENT WOOD IN CONSTRUCTION MATERIAL (BY WEIGHT) , 1951-71
13.6
13.4
13.2
13.0
12.8
12.6
12. 4V H
12.2
12.0
11.8
11.6
11.4
11.2
11.0
10.8
10.6
10.4
10.2
10.0
9.8
9.6
\
\
\
\
\
\
\
X
N
t
1
\
\
\
\
\
\
V / ^^
\ t X.
"*""'ť/ \
\
\
\
i
^" ^\
\
\ ^\
\
^
1 i i -~~i i 1 1 : 1
51 53 55 57 59 61 63 65 67 69 71
Year
Source: See Reference 9
-13-
-------
Figure 7
FREQUENCY DISTRIBUTION OF
PERCENTAGE COMBUSTIBLE DATA
300 .
200
% Range No. of Observations
0-9.9 275
10-19.9 114
20-29.9 113
30-39.9 50
40-49.9 42
50-59.9 42
60-69.9 62
70-79.9 80
80-89.9 79
90-99.9 144
TOTAL
1001
100
10
20
30
40
50
60
70
80
90
100
Estimated % Combustible
-14-
-------
Table 3
NOMINAL HEATING VALUE OF VARIOUS WASTES AND FUELS
Type of Waste or Fuel
Commercial Wastes
Residential Wastes
Industrial Wastes
Wood Wastes
Anthracite Coal
Bituminous Coal
Natural Gas
No. 2 Fuel Oil
* Source: See Reference 7
** Source: See Reference 8
Value (1000's BTU/lb)
8.6*
13.6'
**
**
14.1
1049.6 BTU/ft'
19.9
.**
Table 4
COMPARISON OF ANNUAL BTU POTENTIAL
Pennsylvania Anthracite
Bituminous
Total BTUs from Coal
Demolition § Construction Waste
Forest and Primary and Secondary Wood Industry Waste
% Fuel Value in Demolition & Construction
Compared to Forest and Wood Industry Waste
% Fuel Value in Demolition and Construction
as Compared to Coal
* Source: JACA primary data and calculations
** Source: See Reference 8
*** Source: See Reference 3
BTU x 1014
1.9 **
162.5 **
164.4 **
3.75*
23. ***
1 /* 0, ft
ib-6
-------
ECONOMICS OF WOOD RECOVERY
The supply-demand relationship for waste wood either as a fuel or as
a product is quite elastic because of the ready opportunity for substitutions
of other supplies that can satisfy the same demand. For example, as a fuel,
waste wood would be in price competition with gas, oil, and coal. Virgin
material would compete in paper pulping and construction.
There are three technical factors that affect the demand for waste wood.
These are (1) availability, (2) ability to separate and (3) quality of recover-
able wood fiber.
A predictable supply of wastewood of sufficient quantity to serve a reuse
facility for either demolition or construction waste is an important problem.
It is difficult for any potential reuse operational site to predict the level of
demolition and construction activity over a time period commensurate with a size-
able capital expenditure having an attractive retum-on-investment. The return-
on-investment analysis has two ingredients, the expected rate of return and the
duration of that return. A break even analysis is only meaningful in the context
that the return will continue for some appropriate period after the breakeven
point has been reached. It is during this period, subsequent to the breakeven point.,
that the major returns will be realized.
Demolition and construction activity vary considerably over time, and
geographic locations as indicated by the data on demolition activity and the
information on new construction in Figures 1, 2, 3, and 4. Furthermore the study
indicated that it is difficult to anticipate construction and demolition activity
several years in advance.
Inability to accurately predict the supply of demolition and construction
waste can be translated into the necessity of having large storage facilities to
accommodate supply surges and to provide buffer storage at times of low levels of
supply.
The second problem with supply of demolition and construction waste is that
an added step with respect to wood waste recovery presently being practiced must
be taken. The waste is a mixture of building materials including wood, impregnated
paper, bricks, concrete, metals, glass, plastics, stone and clay products. This
waste must first be separated into its principal components, i.e. wood combusti-
bles, stone, brick, clay or concrete, and metals. Ancillary to this separation
process is the necessity of having the ability to dispose of the waste components}
other than wood}on site. This study indicated that for construction waste an aver-
age of 48% was wood and for demolition waste 39%. Thus the separation process
will result in disposal of large quantities of non wood waste.
The mixture of woods, degree of weathering, moisture content, and the ten-
dency of the wood to retain bits of concrete, soil, brick, etc. poses a quality
-16-
-------
control problem. If the wood is to be used as fuel then the principal restric-
tions are on size and moisture content. High moisture content wood might re-
quire pre-drying. Chips used as pulp in paper operations tend to have more
strict specifications as to appearance and foreign material. The relatively new,
and in most cases experimental use of wood chips in such construction material
as particle board, has not yet generated quality standards. Currently, some
research effort is being directed toward quality standards.
The above three points concerning available supply, need to separate,
and quality control are a strong inhibiting force in the use of demolition and
construction wood wastes. The other side of the problem involves market demand
for recovered wood.
To be economically viable,any conversion to a new product or fuel use of
the wood fraction found in demolition and constructionjinust have long range mar-
kets. At present, there is practically no use made of such wood waste because
of the technical, economic and institutional problems associated with the three
major potential uses of such waste.
Wood Fraction as a Fuel or Fuel Supplement
Demolition disposal operations do not require fuel. Potential users in
the economic area of the disposal site must have special equipment for
utilizing the wood fraction such as:
Conveyors are necessary for feeding wood into the furnace
A furnace for wood combustion is needed
A handling system for ash by-product is necessary
Air pollution control devices are needed.
Wood Fraction as Pulp Fibers
Economics in transportation call for close proximity to user.
Dirt, differences in color or texture cannot be tolerat&d in
some paper plants.
Asphaltic substances may be deleterious to paper production.
-17-
-------
Wood Fraction as a Building Material Substitute
No waste wood product is now used.
Building code requirements must be met when product is
available.
The reluctance to accept new products on the part of building
trades, unions, and other construction industry personnel
must be overcome.
In most instances the market FOB user site for wood from demolition and
construction waste processed at a dump or other central point would have to re-
cover these costs:
Separation from non-wood waste
Disposal of non-wood waste
Reduction in size and screening for use
Pre-drying for some selected uses
Delivery to site
Transportation to user
Profits for operator.
An offset to these costs is the fee for dumping the material, which may
run from $17 to a high of $120 for a 50 cubic yard load. Also as disposal sites
become more scarce, there will be an increase in dumping fees. The break even
equation becomes:
Disposal fee + wood product recovery returns
processing costs
-18-
-------
on an annualized basis:
Disposal fees + Product sales _ Annualized Cost
Year (depreciation,
0 S- M, taxes §
insurance)
PILOT PLANT OPERATION
Two operations were identified and visited during the course of the study
that partially approached an operation for recovery of demolition and construc-
tion waste. Both operations had contracts with one or more paper manufacturers
to deliver wood fibers for pulping operations.
The first system located in San Francisco used a total dry operation to
produce the fibers from wood debris from building, industrial and food processing
industries.
The raw waste, usually picked clean of plastic-sheet and bags which could
clog the screens, is first broken by running over it with a cra\vler-tractor. The
broken waste is then pushed by the tractor into a recessed part of the steel
faced reinforced concrete floor, from which point it enters a hammermill. Large
metal pieces are removed by hand before entering the hammermill. The output from
the haramermill consists of slivers and small chunks of wood together with nails
and other small pieces of metal. The material then moves from the hammermill to
a screen operation by conveyor. Ferrous metal is removed enroute by magnetic
separator and conveyed to a bin. The screening operation consists of two vibratory
screens. Material failing to pass the first 4 inch screen is fed back to the
hammermill for further processing. Material passing both the first and second
screens feeds to a mulch bin for subsequent sale to nurseries. Material of the
right size passing the first, but not the second screen is fed to bins for ship-
ment to a pulp operation.
It is imperative that the material received be practically all wood and
free of plastic, cement bags, etc. In return for bringing only clean wood waste,
the hauler is charged no dumping fee.
A second operation in California utilizes a different system for separating,
size reduction and screening. Dump trucks unload material either directly into a
receiving pit or on a surge pile. A crawler-tractor crushes large scraps into
manageable size and loads them into a 20' deep by 40' long receiving pit which is
equipped with two bottom drag chains. From the drag chain the unprocessed wood
moves up a 6' conveyor to a flotation vat where an agitator knocks off pieces of
metal, gravel, dirt and other material from the floating wood. The damp clean
wood moves from the flotation unit by another 6' conveyor to a wood hog which
utilizes a punch-and-die cutting action to reduce the size of the wood. The
-19-
-------
material is then transported on a 2' belt to two vibrating screens. On the way
a magnet removes the ferrous metal scrap. Fines and sawdust are stored on the
ground for aging prior to sale as mulch. Oversize pieces are sent back to the
hog, and acceptable fibers are sent directly to waiting vans. This system can
handle more non wood constituents than the first system, but in its present con-
figuration could not handle the types of general demolition waste found in most
of the 1001 loads at the 29 dumps visited in this study.
The risks associated with the supply and demand needs for reuse operation
together with the capital needed for development of a system of high utility
for processing general demolition and construction waste without presorting tends
to inhibit development of this market. Government activity to reduce the risk or
capital investment needed would act to spur subsequent private industrial entrance
into this field.
METHODOLOGY & FIELD DATA
This study is characterized by the large amount of primary data that was
gathered through field work, response cards, and interviews. In conjunction with
this work a new technique for determining percent combustibles in mixed bulky
waste was developed and evaluated.
The study results should be viewed in context with the methods used. Be-
cause of the heavy reliance on field-generated data, the collection and analysis
of the data and methods employed will be discussed as follows:
Essentials of the Methodology
Training
Statistical Methodology
Stratifying the Sample
Field Implementation
Density Determination
Volume Determination by Response Cards
Wood Waste Generated by Construction Activity
Estimating the Total Number of Buildings Demolished
and Constructed Per Year in the United States
Estimating the Energy Potential from Demolition and
Construction Waste on National and Local Levels
-20-
-------
Essentials of the Methodology
The primary purpose of the study was to determine the amount of demolition
and construction waste generated in the United States annually and to determine
what fractions of these total figures represent the combustible portion.
At the outset of the study it was intended that the following equation
would be used separately for demolition waste and for construction waste to
arrive at the number of tons of combustible debris generated annually from demoli-
tion and construction activities:
(constructed)
average yd tons buildings demolished % combustible _ tons of combustible
building yd3 year by weight year
CD (2) (3) (4)
(1) was determined from primary data obtained from members of the demoli-
tion and construction industries;
(2) was calculated from the weights of thirty sample truckloads of demo-
lition waste and thirty sample truckloads of construction waste of known volumes;
(3) was determined by extrapolation from the figure on total number of
residential buildings demolished in the United States, found in U. S. Bureau of
the Census Report #C45, to obtain an annual figure on all categories of residential,
commercial, and industrial buildings demolished.
Extrapolation was based on the ratio of residential buildings constructed or
demolished to non-residential buildings (commercial and industrial) constructed
or demolished. Data was derived from the annual building construction and demoli-
tion permit records of the ten cities visited. These calculations are based on
the numbers of buildings demolished and constructed after obtaining a permit and
do not cover unpermitted operations. Unpermitted operations occur most often in
rural areas where levels of demolition activity are low and are expected to repre-
sent a negligible part of the total activity.
(4) was obtained from statistical analysis of the data collected at. disposal
sites in Philadelphia, Los Angeles, Minneapolis, Houston, St. Louis, Miami,
Pittsburgh, Chicago, Atlanta and Detroit. A total of 29 disposal sites were visited
and 1001 truck dumpings observed.
Early in the planning stage of the work it was agreed that "combustibles"
would be construed as wood waste only.
-21-
-------
Training
The data collection methodology for part (4) of the equation used visual
estimation to measure percent combustibles by weight. Therefore, special
training was instituted. The questions to be answered in the training phase
were:
1. What were the best techniques for observing and estimating
the volume of construction and demolition waste in a disposal
operation?
2. What was the accuracy of the estimation?
3. Was the accuracy improved significantly by averaging the
estimates of two or more observations?
A six week training session was initiated during which these issues were
addressed. It was conducted at a suburban Philadelphia disposal site permitted
by the State of Pennsylvania for demolition and construction waste. After meet-
ing with site personnel, several areas were designated and cleared so that the
training waste loads could be kept separate for future analysis. Arrangements
were made with a local NADC demolition contractor to have his incoming 50 yd.^
trucks stop at a nearby scale to obtain gross and tare weight for the vehicle.
Trash bins of 12-18 yd.~* and a dump truck were used in the separation of the
wood fraction. The training area is diagrammed in Figure 8.
When a weighed truck pulled into the landfill area, the observers positio
themselves around the rear of the truck at safe distances and unobstructed obser
tion points. The best location proved to be about 30 from the rear centerline
the truck. Observation of material was made while the load was dumped and after
it was on the ground to make an estimate of the amount of wood and wood products
in the load. Because this was a training exercise, estimates of the first loads
were made on a volume basis until the observers developed a correlation between
the volume and weight percentages.
Estimates of weight, date, time, source, truck weights, and general comme
were recorded on a data sheet. All loads were photographed for subsequent analy
Figures 9 through 11 show some of the loads.
Four JACA employees, went through the training program. Two held Bachelc
degrees and two were technicians with several years experience. It was not knou
how many loads would be required to develop an acceptable degree of accuracy (2C
Progress was monitored and terminated at 32 loads when accuracy was satisfactor>
-22-
-------
O bo f-<
O C O ii
cd -H CL- iI
CX -H
rt
to f-i
O H o
p fn
o tp <
c o
o
c
ID
c
I/) O
t3
rt
C rt X
= o o
x s ^
o
CO
rt
o
f-c
p
H
CO
w
(SO
-------
Figure 9
TYPICAL 40 YD.5 LOAD OF DEMOLITION WASTE
AT LANDFILL
PHOTO A
mounted separately for reproduction
Figure 10
AN UNSEPARATED LOAD OF DEMOLITION WASTE
WITH A HIGH PERCENTAGE OF WOOD
PHOTO B
mounted separately for reproduction
-24-
-------
Figure 11
AMOUNT OF WOOD HAND-SEPARATED FROM DEMOLITION WASTE
IN 1.5 HOURS DURING THE TRAINING PERIOD
PHOTO C
mounted separately for reproduction
-25-
-------
The loads were hand separated into wood and non-wood piles. Portable
trash durapsters and a dump truck were used to transport the separated materials
back to the scale. Because of the size and weight of some of the beams and con-
crete blocks, a front end loader was occasionally used as an aid in loading the
receptable. Since the debris were often very small sized at the bottom of the
pile, and these smaller pieces were often ground up when the front end loader
raked the pile, the weight percentages of the remaining 10% or less were often
estimated and added to the weigh ticket totals. The estimated remainder was less
than two hundred pounds. All of the debris came from the Philadelphia area.
Information as to the type of structure from which it originated was obtained from
the drivers or dispatcher.
The actual percent combustible was obtained in the following manner:
_ gross weight of separated materials - wt. of container ,-
om us i es gross weight of incoming truck - wt. of truck
Thirty-two loads of demolition waste were hand separated and weighed during
the training period. It was noted during the training period that demolition
loads tend to occur most frequently with high percentage combustibles or low per-
centage combustibles rather than in the 30-70% range (Figure 12). This same con-
dition prevailed in 1001 observation in later field work (Figure 7).
A summary of the training results appears on Table 5. This summary shows
an error of 5.49% when averaged over all 32 loads. This was within the pre-deter-
mined 20 percent tolerance established for the practical utility of the visual
estimating technique. Table 5 also shows a reduction in error as the training
continued. The error of the average of the first 10 loads was 4.99% and of the
last 22 loads 1.75%. The average of trainees' absolute error per training truck-
load is shown in Figure 13.
Statistical Methodology
Predictive Versus Empirical Approaches
There are two techniques for estimating the quantity of demolition and con-
struction waste and the percentage that is combustible. In the case of demolition
waste, predictive techniques involve determining the composition of buildings pre-
sently being demolished on the basis of knowledge of materials and quantities,
as well as the original construction techniques. Information concerning the age
-26-
-------
Figure 12
FREQUENCY DISTRIBUTION
OF ACTUAL COMBUSTIBLE PERCENTAGE
-LOADS 1 to 32-
o
bo
C
0)
W)
c
o
o
f-4
o
o.
cd
to
C
H
rt
O
E
A
10"
10 20 30 40 50 60 70 80
Actual Combustible Percentage
90
100
-27-
-------
Table 5
SUMMARY OF TRAINING RESULTS
A. Total of 32 Loads
Actual % Combustible 46.09
Estimated % Combustible (ave.) 51.58
Error 5.49
.... c_^--_ +.-- f -i-i T-n Observer Observer Observer Observer
1. Individual Estimates (all 32)
Estimated % Combustible 50.20 57.36 44.32 54.45
Error 4.11 11.27 -1.77 8.36
B. Last 22 Loads
Actual % Combustible 44.21
Estimated % Combustible (ave.) 45.96
Error 1.75
1. Individual Estimate (last 22) #1 #2 #3 #4
Estimated % Combustible 52.2 53.2 40.8 54.4
Error 8.0 8.99 -3.37 10.19
C. First 10 Loads
Actual % Combustible 50.21
Estimated % Combustible (ave.) 55.20
Error 4.99
1. Individual Estimates #1 #2 #3 Ł4
Estimated % Combustible 46.17 66.5 51.67
Error -4.04 16.29 1.46
-28-
-------
0>
W3
H
C_J
ci
oi
-29-
-------
of a structure being demolished, as well as the number of units of each type
provides an estimate of the amount of demolition waste generated as well as
the percentage that is combustible. This estimate is employed in conjunction
with information that involves the use of predictive techniques covering the
number of buildings demolished, the average number of cubic yards of waste
generated per building and the density of the waste per cubic yard, to yield
the total number of tons of combustibles available.
There are limitations that affect both techniques. Those affecting
the predictive technique have been summarized in a report written by Wiesman
and Wilson . First, estimates of building composition at demolition will incor-
porate errors, since alterations, additions and replacements take place after
the building is constructed. Second, the set of buildings being demolished is
not a homogeneous combination of all types and sizes, and the use of the computed
content of the "average" building may differ significantly from those being razed.
Third, clean fill, or, waste not containing wood, metals, plastics, and etc., may
be short circuited and never reach the dump site. Fourth, on many jobs, the
demolition contractor leaves any basement or foundation one foot below grade in-
tact, and fills the basement with non-wood waste from the structure. Any waste
so diverted is not truly available, but is part of predictive estimates. Finally,
original construction records kept by local governments often lack detail and
inaccurate estimates may result.
Empirical sampling, combined with predictive techniques has some limitations
First, sampling involves the random selection of sites and times of observations
and, it is often not possible to achieve this randomness because of time and
budget constraints. To the extent that randomness is lacking, errors may enter
into the estimation process. Second, the type of random sampling employed in this
study involved a process of observation by trained technicians, and thus is sub-
ject to errors of measurement. Finally, since some predictive techniques are
employed, these estimates may incorporate some of the inaccuracies mentioned above.
The major sources of possible error involve the use of cubic yards of waste gen-
erated per demolition and the total number of buildings being demolished. To the
extent that these averages are not representative, estimates will be off. Extreme
caution was used by the contractor in the careful inclusion of randomness in the
selection process and through training of the technicians in an attempt to minimize
the first two problems. The third problem is reduced through gathering of back-
ground and on-site data and careful calculation of averages.
Errors in Sampling
The sampling process is subject to two types of errors. The measurement
or non-sampling error occurs because of a difference beteen the actual value and
the measured value. This kind of error arises from factors such as imperfect
-30-
-------
observation, faulty questionnaires, inaccurate tallying, etc. The sampling
error results from the chance selection of sampling units. This error
occurs when a partial observation of the universe takes place. If the entire
universe were studied, the sampling error would be zero.
The total error in a statistical survey such as this one is the sum of
the measurement error and the sampling error. The major concern is to minimize
the total error. Reduction of the measurement error is achieved by defining
precisely the population to be studied and its traits, by refining the measure-
ment process to the highest degree, and by training the individuals doing the
measurement as thoroughly as practicable. These precautionary measures are
usually costly, leading to the necessity of using small samples. However, small
samples tend to have larger sampling errors. Therefore, there are two options
available: the sample size can be kept small and the measurement made in a
sophisticated manner, or the sample size can be large and the measurement made in
an unsophisticated manner. The following example should help illustrate a method
of estimating the percentage of combustibles in construction and demolition waste.
Assume that a sophisticated measurement process would limit the sample size to
50 observations or an unsophisticated technique would employ trained observers
to estimate the combustible proportion and would allow for a sample size of 300
observation's. Assume that the measurement error in the first case would be five
percent and in the second case it would be ten percent. In estimating proportions,
the formula for the maximum sampling error is given by:
a,
5 = sampling
z - normal curve deviate which is determined by the
level of confidence
IT = universe proportion being estimated
In the absence of information concerning the possible size of ir, it is
assigned a value of 1/2, which maximizes the expression TT(I-TT), and therefore,
maximizes 6 for a given level of confidence and sample size. In the cases cited
above, if a level of confidence of 90% is employed:
m K - i A/K /(.5)(i-.5) = .116
(2)
-------
Using a level of confidence of 95%, we observe:
x i n^ /(.5)(1-.S) - .056
695 = l'96 I 300
V
Since:
Total Error = Sampling Error + Measurement Error, the following
results are obtained, with 90% confidence:
Case 1 Total Error = .116 + .05 = .165
Case 2 Total Error = .047 + .10 = .147
with 95% confidence:
Case 1 Total Error = .139 + .05 = .189
Case 2 Total Error = .056 + .10 = .156
At both levels of confidence, the total error is smaller when the sample
size is increased at the expense of a larger measurement error. In the case of
this study, we chose the option of increasing the sample size as opposed to re-
ducing the measurement error, and the cost of doing so was significantly less
than reducing the measurement error.
Stratifying the Sample
In stratified random sampling, the universe is classified into mutually
exclusive subgroups or "strata", and samples are drawn from each of these strata.
-32-
-------
Sample statistics are calculated from each of these strata and are combined to
yield an overall estimate of a population parameter. The basic purposes of
stratified, as compared to simple random sampling, are to obtain a sample that
closely resembles the universe from which it was drawn and to reduce sampling
errors. These objectives are accomplished by grouping together into strata those
elements which are more alike with respect to the characteristic under inves-
tigation than are elements in the universe as a whole. Stratification is most
effective when the elements within strata are as homogeneous as possible, as
regards the property to be studied, and the differences among strata are as great
as possible.
In this study the demolition samples were to be stratified by type of
structure. Initially, four strata were contemplated: residential, multi-unit
residential, commercial and industrial buildings. Experience in the field led
to a reduction in the number of strata to two: residential and other. This was
due to the fact that many cities did not distinguish among the various types of
structures or the required number of observations to define strata was not gen-
erated in each of the four areas. The same strata were initially proposed for the
construction data. The number here was also reduced to two for the same reasons.
The stratification by phase of construction was also considered but was abandoned
after attempts to gather the data proved highly impractical within budget and time
limitations.
Site Selection
Figure 14 indicates how the geographical site selection was accomplished.
The United States was broken down into four areas, and a percentage of the total
population calculated for each area. The ten cities were assigned to an area
based on population percentages. This technique resulted in a selection of cities
that would represent any regional peculiarities of building material and techniques,
Ten cities were selected on this area breakdown on the basis of construc-
tion and demolition activity. Within each of these ten cities disposal sites
were chosen on the basis of activity at the site and the willingness of the opera-
tor to cooperate by allowing field technicians to make on site observations.
Stratification by Building Type
Once the cities had been selected, the desired number of observations
in each stratum (residential and other) for construction and demolition were
determined by averaging the number of construction and demolition permits issued
in each city during the period 1970-1975. These permits were used to establish
the ratio of residential to other units constructed or demolished. This ratio
was then applied to the number 30, the desired number of observations
-33-
-------
(-1
3
bo
H
, u
-34-
-------
for each city, to determine the desired number of observations in each of the
two strata for that city. The desired and observed stratification for demoli-
tion and construction are shown in Table 6.
Teams of two technicians each were sent to disposal sites in each of the
ten cities. Each technician independently observed loads as they were dumped
and the individual readings were averaged to obtain a better estimate (smaller
variance) of the percent combustible. Varying lengths of time were spent in an
effort to gather enough data to meet the desired stratification, which was
universally met.
Since the observed samples exceeded the required or "desired" number at
each location, a weighting technique was devised which allowed the use of all
the data generated, even where the actual number observed exceeded the desired
number. Weights were assigned in such a way as to keep the strata in the proper
desired ratio.
The weighted technique calculated the percentage combustible for each city
as the weighted mean of the two strata, employing the desired number of observa-
tions in each strata as weights. An example of this technique for Philadelphia
is shown on Table 7.
To determine the overall percentage combustibles, the arithmetic mean and
standard deviation of the ten weighted percent combustibles was calculated. Em-
ploying this datum a confidence interval for the overall percentage of combusti-
bles was calculated as follows:
95%
<_Mp <_Xp+t (s/n)
where :
Xp = mean of the ten weighted percent combustibles
s = standard deviation of the ten weighted percent combustibles
t ^ = coefficient which is determined by the level of confidence
n = sample size (10)
Mp = mean percent combustible for the universe
-35-
-------
Table 6
DEMOLITION AND CONSTRUCTION FIELD DATA STRATIFICATION
Demolition
Construction
Residential Non-residential Residential
Philadelphia
Los Angeles
Chicago
Detroit
Miami
Houston
St. Louis
Atlanta
Pittsburgh
Minneapolis
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
Desired
Observed
25
39
24
26
22
35
25
36
26
56
24
26
22
23
26
57
26
57
24
35
5
66
6
20
8
55
5
6
4
12
6
10
8
25
4
5
4
11
6
18
25
26
23
34
26
26
25
31
20
22
26
38
17
21
19
20
15
15
17
20
Non -residential
5
22
7
14
4
10
5
11
10
10
4
8
13
13
11
13
15
15
13
13
-36-
-------
Table 7
CALCULATION OF PERCENT COMBUSTIBLE DEMOLITION WASTE, PHILADELPHIA, PA
Number of Observations
_ . j _., , Percent
Deslred Observed Combustible
Residential 25 17 45.1
Non-residential 5 21 20.7
Weighted percent combustible = R, P + N, P
or a n
Rd + Nd
where:
R, = number of residential observations desired
d
P = percent combustible observed from residential readings
N, = number of non-residential observations desired
d
P - percent combustible observed from non-residential readings
Weighted percent combustible = 25 (45.1) + 5 (20.7) .. _.
= 41.04-6
-37-
-------
For demolition s = 12.12
38.60 - 2.262 (^|4r3 1 Mp <_ 38.60 + 2.262
29.92 <_ Mp < 47.28
For construction: s = 8.34
47.59 - 2.262 ( |^|) < Mp < 47.59 + 2.262 (
41.62 Ł Mp Ł 53.56
The interpretation of the confidence interval is as follows: there is
a ninety-five percent certainty that the mean percent combustibles of all demo-
lition waste is between 29.92% and 47.28%. The calculation of the percent
combustible of construction waste and the interpretation of the resulting con-
fidence interval is analogous to that for demolition. Once the mean percent
combustible was determined, the total number of tons of combustibles was calcu-
lated by employing this percentage in conjunction with data developed on the
basis of predictive techniques. The calculation for demolition waste is:
TCD = A D
where
TCD = total number of tons of combustibles per year from demolition
waste
A = average number of cubic yards of waste per building
D = average density of demolition waste
B = number of buildings demolished per year
P = percent combustible of demolition waste
-38-
-------
The average number of cubic yards per building was determined from an
NADC cooperative questionnaire. The average density was determined in field
experiments with JACA technicians. The number of buildings demolished per
year was determined from literature searches. The percent combustible was
calculated above.
In the case of construction waste it was impossible to develop data
from haulers as to the average volume of loads from typical waste sites, as was
done via the questionnaire for demolition waste. These data were not available
because of the long duration of construction activity.
The calculation for construction waste differs significantly from that
for demolition waste.
TCC = P -W
where:
TCC = total number of tons of combustibles per year from construction
waste
P = percent of wood that is wasted in construction
W = total amount of wood consumed in building and construction.
The percent of wood that is wasted in construction was determined from
information gathered through literature searches and discussions with building
contractors based on their estimating procedures. The total amount of wood con-
sumed is taken from existing government sources.
Collecting Field Data
The primary segment of this study utilized field measurements of properly
stratified samples. The plan was to use a total of at least 600 random samples
-39-
-------
from three sites at each of ten cities visited. In the final analysis, 1001
samples were obtained. This sample size was large enough to fully account for
geographical differences in building sites and to provide for different sizes
and types of buildings. Trained observers were used to estimate the percent
combustibles as the trucks dumped their loads. A satisfactory measurement
accuracy by this means would be 20% for all sample totals. The training showed
that much better than 20% accuracy was obtained. The objective in each city
studied was to collect enough samples (percent combustible values) for demoli-
tion and construction waste to meet the previously determined requirements in
each category of a sixty sample, stratified set.
The first data collection was done in the Philadelphia area. The Annual
Building Construction Reports for the years 1970-1975 were obtained to determine
the stratification, in preparation for actual fieldwork. Construction and
demolition permits for the past 5 months were examined to determine the recent
activity levels in each stratum. The contractors whose names appeared frequently
on these permits were contacted to determine which Philadelphia area landfill
sites they were using and arrangements were made with landfill operators to
station field observers at five local sites.
The activity in demolition and construction was extremely low during the
initial data collection phase as compared to the activity levels during the
earlier training session. This indicated that the generation of demolition and
construction waste was sporadic. As a result, the fieldwork was rescheduled to
coincide with periods of high demolition and construction activity in order to
get the most samples during field visitations.
It was evident that a refinement in the scheduling of cities was extremely
important. Following discussions with the National Association of Demolition
Contractors of our objectives. Mr. Ron Dokell, President, offered the Association's
assistance and, together with the Energy and Recycling Committee provided assis-
tance pertaining to our visits to the ten cities. The NADC committee supplied
names of local members who aided in scheduling of fieldwork to coincide with
high demolition activity. These contacts were also helpful in directing us
to the landfills where most of the waste was being hauled.
Arrangements were made to visit three landfills at each of the ten cities.
Because one of the determining factors as to where demolition and construction
will be dumped is the cost of hauling based on distance, the observation of
three local sites helped neutralize intracity peculiarities in activity or com-
position of waste.
The method of obtaining percent combustible data followed the procedure
described in the training section. At safe observation points, each load of
construction and demolition debris was analyzed while being dumped and while
on the ground to estimate the percent combustible by weight. The driver of
the vehicle was questioned briefly as to what building category provided the
waste.
-40-
-------
Depending on activity levels in each city, observers spent one to two
weeks on site, and in some instances had to return to a city to satisfy the
stratification requirements. The stratification for the city was referred to
periodically during the week to determine whether sufficient residential,
commercial, and industrial loads were being sampled, Photographs were taken
at several dumping sites for future reference.
Construction activity evidenced at the ten cities was mainly in the form
of small truckloads and portable bins of wastes generated in renovation, roof-
ing, and siding projects. Very little waste was observed from new building con-
struction. One explanation for this phenomenon is that waste generated on con-
struction projects is disposed of on site rather than paying dumping fees at a
landfill. Therefore, it was necessary to schedule return visits to obtain
fully stratified, 60 sample sets of construction samples.
Some of the city building reports did not distinguish between commercial
and industrial demolitions. Therefore, these two categories were combined into
"non-residential" construction and demolition for all the cities surveyed. De-
creasing the number of strata did alleviate some of the problems in obtaining a
proper sample. After the data on the 10 cities was reviewed, it was necessary
to return to Miami, St. Louis, and Chicago for sufficient data on construction.
Sufficient demolition data was collected in all ten cities on the first visit
because the demolition volume to building volume ratio is so large that contin-
uous landfill dumping is necessary.
Individual observers' estimates were averaged at each of the ten cities,
and it was found that the difference between one observer's average for the
entire sample and the average of two observers for the entire sample was negli-
gible. Therefore, one observer was sent to each of the four cities where insuffi-
cient construction waste data had been collected on the first visit.
The field data results collected on percent combustibles in the ten cities
weighted in accordance with the method shown in Table 7 are tabulated in Table 8.
Density Determination
The density of the loads was a second piece of primary data to be obtained
in the field. During the initial training session, thirty-two loads of demolition
waste were weighed and separated at a local landfill site. The demolition trucks
were 50 cubic yard size capacity.
In order to use the equation presented in the section entitled National and
Area Estimates, a value for the average density of demolition waste was needed.
The concentration of the various components of demolition waste such as wood,
brick, concrete, dirt, and the permitted road weight of a particular truckload
help determine the density of the load and the volume to which the truck can be
-41-
-------
Table 8
City
Philadelphia
Los Angeles
Chicago
Detroit
Houston
St. Louis
Miami
Pittsburgh
Atlanta
Minneapolis
WEIGHTED PERCENT OF WOOD
IN DEMOLITION AND CONSTRUCTION
-FIELD
DATA-
% Wood
Demolition
41
63
44
42
27
51
37
27
20
34
WASTE
% Wood
Construction
42
48
60
43
63
50
43
36
40
51
Average Weighted % Combustible
39
48
-42-
-------
r
filled. Often a 50 yd. capacity demolition truck with high density components
may be filled to only 20 yd.3 whereas a truck with a high percentage of
wood is generally filled to volume capacity. The average volume of thirty-two
50 yd truckloads averaged 40 yd.3.
weight of load
load ~ 1080 ft-5
When the average densities for the 32 loads were calculated using the
above equation, the value for the density of demolition waste was found to be
25 lbs/ft.3. This figure was considered accurate for this study because the
percentage combustible values for the 32 training loads had good distributive
representation from both low percentage and high percentage combustible loads.
Volume Determination by Response Cards
In order to use the equation presented in the National and Area Estimates
section a value for the average number of cubic yards of waste generated per type
of building demolished or constructed was needed.
3
Avg. yd. x tons bldgs. (const./demo.) % combust. _ tons of combust,
Bldg. (const./demo.) yd.^ year X by weight ~ year
To determine the volume of waste generated in the demolition of buildings,
the NADC cooperated with JACA Corp. in administering a response card program,
which supplied data collected by seventeen volunteer members of NADC. Informa-
tion on the response card is shown by the following sample.
Respondent No. 20
1) Job Started 8/9 2) Job Finished 8/12
Enter No. of Enter
Check Type of Building Truckloads Truck Size
3) Residential 4) units = 2_ 5) 24 6) 40 cu.yds.
7) Commercial
8) Industrial
-43-
-------
Respondents remained anonymous through a respondent numbering system,
and the geographical distribution was statistically sound. One card was com-
pleted per demolition job. The information entered on the card included dura-
tion of job, type of building, number of units if residential, number of loads,
and the truck size. The sample card above indicates 960 cubic yards of demoli-
tion waste were generated from the duplex structure in 4 days.
The NADC response card program ran for approximately three months, yield-
ing 200 responses from the seventeen respondents. Following completion, cards
were separated according to building category, and the average volume of waste
per demolished building was calculated for residential, commercial, and industrial
buildings.
AVERAGE VOLUME OF WASTE PER BUILDING
Residential 450 yd3/bldg.
Commercial 2022 yd /bldg.
Industrial 3860 yd3/bldg.
Overall 1370 yd3/bldg.
From other information included on the response cards, the average duration
of a demolition job could also be determined.
AVERAGE DURATION OF DEMOLITION JOB
Residential 3.87 days
Commercial 9.56 days
Industrial 14.7 days
Overall 6.92 days
The distribution of response cards by respondent and the cities represented
by the response program are shown in Figure 15 and Table 9 respectively.
-44-
-------
Figure 15
NUMBER OF RESPONSE CARDS COMPLETED BY RESPONDENTS
espondent
No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17 -f
18
19
20
21
22
23
24
25
-45-
-------
Table 9
CITIES REPRESENTED IN VOLUNTEER RESPONSE CARD PROGRAM
Skippack, Pa. (Philadelphia)
McKeesport, Pennsylvania
Windsor, Connecticut
Atlanta, Georgia
Akron, Ohio
Columbus, Ohio
Fort Wayne, Indiana
Chicago, Illinois
Lyons, Illinois
Riverside, Illinois
South Suburban, Illinois
Maywood, Illinois
Elmhurst, Illinois
Skokie, Illinois
Detroit, Michigan
St. Paul, Minnesota
Houston, Texas
Tucson, Arizona
Seattle, Washington
-46-
-------
Wood Waste Generated by Construction Activity
As previously discussed, lengthy job duration, variation in methods
of disposal of construction waste, and the inability of members of the
hauling industry to provide average estimates of volume of waste generated
per each construction job frustrated attempts to determine the volume of wood
waste generated in the construction of a residential building, a commercial
building, and an industrial building. Therefore, it became necessary to develop
data on the basis of estimates made by construction contractors on the percentage
of wood waste generated on their jobs. Contractors were requested to estimate
the wood waste as a percentage of wood ordered per construction job. The
results of this survey of 20 contractors indicated that an average of 7.4%
of wood delivered is wasted in the process of construction. This average was
tested for significance by the following equation:
At a 95% level of confidence,5.28 < y < 9.52
Estimating the Total Number of Buildings Demolished Per Year And Constructed
Per Year in the United States
Annually, the U. S. Department of Commerce, Bureau of the Census issues a
report entitled Housing Units Authorized for Demolition in Permit-Issuing Places.
The report gives the total number of permits issued for housing unit demolition
annually in the 327 U. S. Cities having a population of 50,000 or more. The re-
port also gives a United States total figure based on reports from permit issuing
places authorizing the demolition of 1 or more housing units.
Using the figures presented for the number of housing units demolished in
cities over 50,000 population ( .34 of total U. S. Population 1970 Census) as well
as the individual cities populations, a statistical evaluation was conducted to
determine whether a linear relation existed between population and number of
units demolished using methods of linear regression analysis. The equation
y = -3.09 + .0018 x was derived where y = number of units demolished and x = the
size of population. This equation explained 61% of the variation for this rela-
tion. Therefore, there would be a certain degree of error in pursuing the calcu-
lation of a U. S. total for demolition of housing units by this method.
-47-
-------
While the above equation might be useful in predicting the amount of
demolition for an area, given its future population estimate a more reliable
estimate was needed. Therefore, it was decided that the figures for U. S.
total number of housing units demolished based on reports from permit issuing
places authorizing the demolition of 1 or more housing units would be used.
The Bureau of the Census states that these annual figures are based on reports
from areas which represent about 80-85% of the population and would therefore
imply that these figures probably represent 95% of the total U. S, housing
unit demolition rate.
To convert from units demolished per year to residential buildings
demolished per year, the average number of units per building demolished from
the building reports of the ten cities visited was calculated. This was found
to be 1.4 units per building.
1975 1974 1973 1972 1971 1970
121,972 139,450 149,141 141,915 137,426 122,773
121,972 units/yr - 1.4 units/building = 87,123 buildings/yr.
To determine the total residential and non-residential buildings demolished
per year, the ratio of residential to non-residential buildings,demolished per
year in the ten cities surveyed,was computed and averaged from their respective
building construction and demolition reports. The following equation was used:
, rio. of residential and non-
no, of residential buildings demolished _ residential buildings demolished
year -803 year
87,123 x -~ = 108,497
To determine the total number of buildings constructed per year in the
United States a similar approach was used. Using data on residential construc-
tion obtained from the National Association of Homebuilders which showed housing
units constructed per year (Table 2) the following procedure was used.
-48-
-------
residential units constructed _ av_g/.. units _ residential bldgs. constructed
year ' bldg. constructed year
1,512,900
1 '
residential and non-residential
residential buildings constructed 1 _ buildings constructed
year .724 year
605,160 x ~~ = 835,856
Estimating the Energy Potential from Demolition and Construction
Waste on National and Local Levels
The heating value of wood was taken as 8613 BTUs per pound. By
multiplying the tons of combustibles generated annually from these two
sources, the energy potential of the waste wood can be calculated on a
national level:
Demolition
avg. yd tons bldgs. demo. % combustible BTLP_s _ Total BTU's
bldg. yd0 year ton ~ year
1369.8 x .337 x 108,500 x .386 x (1.72 x 10?) = 3.3 x 1014BTUs
year
Construction
tons of wood consumed annually average % wasted per BTU's _ total BTU's
by construction industry job in construction ton ~ year
35,000,000 x .074 x (1.72 x 10?) = 4.5 x 1Q15 BTU's
year
14
Total energy potential from demolition and construction nationwide=3.75x10 BTU's
year
-49-
-------
To determine the energy potential from demolition waste for a given
local area (i.e. the ten cities included in this study) the above equation
is used with a substitution of number of buildings demolished per year within
the city for the number of buildings demolished on the national level. The
local weighted percent combustible figure (Table 8) can be substituted for
the national average of 39%. The volume per building must be recalculated
in accordance with the average ratio of residential, commercial, and industrial
buildings indicated in the building reports and the appropriate volumes shown
per building.
To calculate the energy potential available from construction waste
on a local level, the amount of wood consumed must be estimated. The
method used was to set up an equality which relates amount of wood con-
sumed by the construction industry on a given level to the number of build-
ings constructed on that level.
Wood consumption Wood consumption
nationwide by by city by construc-
construction industry _ tion industry
buildings constructed buildings constructed
nationwide by city
e.g. Chicago 55,000,000 tons
835,856 bldgs. ~ 2482
x = wood consumption = 104,000 tons per year
by Chicago con-
struction industry
Using the construction waste energy potential equation:
7 1 ^ v 1 fl RTF!
104,000 x .074 x (1.72 x 10 ) = '
year
The energy potentials from both construction and demolition debris calculated
by these methods for each of the ten cities are shown in Table 10.
-50-
-------
Table 10
ANNUAL BTU POTENTIAL OF WOOD WASTE FROM
DEMOLITION AND CONSTRUCTION WASTE FOR
TEN CITIES
City
Philadelphia
Los Angeles
Chicago
Houston
Detroit
Miami
St. Louis
Atlanta
Pittsburgh
Minneapolis
Tons of Combustibles from Demolition
and Construction per Year
1.9 x 105
6.7 x 105
4.0 x 105
1.1 x 105
4.9 x 105
.67 x 105
2.3 x 105
, -42 x 105
1.0 x 105
.67 x 105
BTU's per Year
from Construction
and Demolition
3.3 x 10
12
.11 x 10
12
6.9 x 10
12
1.9 x 10
12
5.5 x 10
12
1.1 x 10
12
4.0 x 10
12
.72 x 10
12
1.8 x 10
12
1.2 x 10
12
-51-
-------
DISCUSSION OF ACCURACY OF RESULTS
This section will discuss errors incorporated in the estimations
of combustibles. Because of the different estimating methods used, for
demolition and for construction wastes, error will be analyzed separately.
Demolition Waste
avg. yd tons buildings demolished % combustible _ tons combustible
year by weight year
(3) (4)
The above equation previously presented was used to develop a figure
for the total number of tons of combustible demolition waste generated annually.
For ease of explanation., numbers are used to label each step.
In the calculation of Step 1, the waste volume data from the NADC
response card program was averaged by building category and the overall
weighted average was then computed. Computations were made for standard
error on the mean for each category as follows:
Residential Waste Volumes
Classes
0-499 66
500-999 27
1000-1499 2
1500-1999 8
2000-2499 2
2500-2999 0
3000-3499 1
3500-3999 0
4000-4499 1
107
The standard deviation of the sample = 658.02
The standard error (SŁ) = 5 = 63.3
~
-52-
-------
Commercial Waste Volumes
Classes
0-1999 51
2000-3999 11
4000-5999 5
6000-7999 2
8000-9999 0
10000-11999 2
12000-13999 0
14000-15999 0
16000-17999 1
72
The standard deviation of the sample = 2756.55
The standard error (S ) S = 324.3
Industrial Waste Volumes
Classes
0-2999 13
3000-5999 5
6000-8999 1
9000-11999 1
12000-14999 0
15000-17999 0
18000-20999 0
21000-23999 0
24000-26999 0
27000-29999 0
30,000 1
21
The standard deviation of sample = 2414.24
The standard error (S ) = S = 290.87
-53-
-------
Residential
Commercial
Industrial
Standard Error
Per Category
63.3
324.3
290.87
Weighting
Factor
.18
.52
.30
Weighted Standard
Error Squared
129.82
28438.10
7614.48
1.00
36182.40
Overall SŁ = / 36182.4 = 190.22
Percent Error - 190.22 = 13.9% - total error involved in the calculation
of Step 1.
In the calculation of Step 2, the weights of 30 loads of demolition
debris, each with a volume of 40 cubic yards, were used to determine an
average density value.
Classes
12000-14999
15000-17999
18000-20999
21000-23999
24000-26999
27000-29999
30000-32999
33999-35999
36000-38999
39000-41999
42000-44999
45000-47999
48000-50999
5
2
6
5
I
4
4
0
1
1
0
1
1
31
The standard deviation of the sample = 9477.64
The standard error (S ) = S_ = 1692.44
Error per ft =
Percent Error =
1692.44
1080
1.57
25
/n
- =1.57
= 6.3%
-54-
-------
Step 3 of the equation shows the number of buildings demolished per
year. Data from the Bureau of the Census Report C45, were used here which
gave the estimated total number of permitted building demolition from 85%
of the population. The error on this part of the equation was assumed mini-
mal since the remaining 15% of the population was probably from rural areas
in which little demolition activity occurs.
Step 4 of the equation is the percentage combustibles by weight data
which was collected during the fieldwork in ten cities. Results of the
training period showed the error to be about 5.5%.
Calculation of total error for demolition (TE) = f E + E + E + E ]'S
v 1 2 3 4
= (13.92 + 6.32 + 0 + S.S2)'1
=(193.21 + 39.69 + 0 + 30.25)^
= 16.22%
and if Step 3 involves a maximum
error of about 5% then,
TE = (13.92 + 6.32 + 5.02 + 5.52)^
= (193.21 + 39.69 + 25.0 + 30.25)^
= 16.97%
Construction Waste
% of wood waste
in construction
tons of wood consumed
annually by construction
Tons of wood wasted by
construction
The above equation was used to calculate the amount of wood wasted in
building construction annually in the U. S, The calculation of the figure for
percentage of wood wasted in construction is the only part of the equation
that involves some error as the estimates of twenty construction contractors
were used.
-55-
-------
Estimated Construction Wood Waste per Construction Job
10
4
10
1
12.
10
10
0
3
10
10
5
1
2
10
5
15
10
15
5
Total 148.5
Average (x) = 7.4 per construction job
The standard deviation of the sample = 4.64
The standard error (S ) = S = 4.64
l
The error (E) = E = 1-° = 13-55
- 7.4
-56-
-------
REFERENCES
1. Chatterjee, Samar, Predictive Criteria for Construction/Demolition
Solid Waste Management. Batelle Columbus Laboratories, Columbus,
Ohio] November 8"," 1974.
2. Wilson, David Gordon, An Investigation of the Potential for Resource
Recovery from Demolition Wastes. Massachusetts Institute of Technology,
July 1976.
3. The Outlook for Timber in the United States, U.S. Dept. of Agriculture,
Forest Service, Forest Resource Report No. 20, 1973.
4. University of Wisconsin, "Trash for Livestock Bedding Feasible." News.
University of Wisconsin Ext. Col. of Agricultural and Life Sciences,
Madison, Wisconsin, 1974.
5. Barkson, Coorts, Roth, "Hardwood Bark Chips", Illinois Parks § Recreation.
November/December 1971.
6. "A New Dimension in Total Tree Chipping", What's New In Total Chip
Harvesting, Vol. 1 No. 5 Morbark Industries.
7. Metcalf and Eddy, Inc., Generation of Steam from Solid Wastes, Environ-
mental Protection Agency Publication (SW-49cl), 1972.
8. Perry and Chilton3 Chemical Engineers Handbook, McGraw Hill Chemical
Engineering Series, 1973.
-57-
------- |