EPA/530/SW-16
OCTOBER 1975
aste i"1
t
-------
An environmental protection publication in the solid waste management
series (SW-163). Mention of commercial products does not constitute
endorsement by the U.S. Government. Editing and technical content of
this report are the responsibility of the Resource Recovery Division
of the Office of Solid Waste Management Programs.
Single copies of this publication are available from Solid Waste
Information, U.S. Environmental Protection Agency, Cincinnati, Ohio
45268.
-------
RESOURCE RECOVERY PLANT COST ESTIMATES:
A COMPARATIVE EVALUATION OF FOUR RECENT
DRY-SHREDDING DESIGNS
Environmental Protection Publication SW-163
in the Solid Waste Management Series
by
Frank A. Smith
U.S. ENVIRONMENTAL PROTECTION AGENCY
July 1975
-------
CONTENTS
Page
List of Tables iv
INTRODUCTION 1
GENERAL METHODS AND DESIGN ASSUMPTIONS 2
What the Data Represent 2
Standardizing the Plant Designs 4
Normalizing the Cost and Revenue Estimates 5
COMPARATIVE SUMMARY OF NORMALIZED CAPITAL
INVESTMENT COST ESTIMATES 10
Total Capital Cost 10
Annualized Capital Cost 10
Capital Cost Per Ton 12
COMPARATIVE SUMMARY OF NORMALIZED O&M COST ESTIMATES 12
SUMMARY OF TOTAL AND NET COST ESTIMATES 14
Total Cost Estimates 14
Net Revenue or Cost Results 17
SUMMARY AND CONCLUSIONS 18
References 20
-------
LIST OF TABLES
TABLE 1 Net Prices Received Per Ton of Product and 9
Revenue Per Ton of Raw Waste Processed: "High"
and "Low" Estimates (1974)
TABLE 2 Normalized Capital Investment Cost Estimates 11
for Four Ury-shredded-fuel Processing Plant
designs
TABLE 3 Normalized Operating and Maintenance Cost 13
Estimates for Four Dry-shredded-fuel Processing
Plant Designs
TABLE 4 Summary of Normalized Cost Estimates for Four 15
Dry-shredded-fuel Processing Plant Designs (Dollars
Per Ton of Raw Waste Input, 1974 Cost Base)
TABLE 5 Summary of Alternative Net Revenue (Cost) 16
Calculations for Four Preliminary Plant Designs
at Two Alternative Capacity Utilization Rates
(Dollars Per Ton of Raw Waste Input, 1974 Cost Basis)
IV
-------
RESOURCE RECOVERY PLANT COST ESTIMATES: A COMPARATIVE
EVALUATION OF FOUR RECENT DRY-SHREDDING DESIGNS
INTRODUCTION
The recovery of salable material resources from mixed municipal
solid waste will generally involve rather complex processing of the raw
collected refuse in large, capital-intensive facilities. Although local
decisions to implement these large-scale resource recovery plants will
probably not be based solely on direct cost considerations, cost comparisons
among alternative recovery options as well as between recovery options
and conventional waste disposal methods will play an important, if not
decisive, role in most community decisions. Resource recovery processing
costs will also be a factor at the State and Federal levels of policy
formulation. For these and other reasons, it is important that sound
data be available in a form useful for comparing alternative projects
and design concepts.
Unfortunately, little useful cost information is currently avail-
able. As of mid-1975, no full-scale mixed-waste separation plants have
been constructed or operated. In the absence of operating data, cost
projections must be based on preliminary design cost estimates derived
largely from experience with pilot-scale operations and equipment supplier
quotations.* Aside from this unavoidable factor, the wide diversity of
competing systems and the different methods of cost-accounting and
estimating used by different designers make relevant comparisons extremely
difficult. In addition, most estimates have been site-specific, reflecting
economic factors such as labor rates, operating schedules, and other
costing parameters peculiar to local circumstances. Thus, even when
available, cost estimates have lacked comparability.
This paper reports the findings of a recent EPA investigation
designed to clarify the present state of knowledge about the cost of
large-scale, mixed-waste processing plants.1 Its more narrow objective
was to provide more definitive comparative cost estimates for one
particular type of mixed-waste processing technology, namely: the
production of supplemental boiler fuel via mechanical shredding and air
*For the U.S. Environmental Protection Agency's most recent status
report on the planning of these facilities, see: Hopper, R. E. A
nationwide survey of resource recovery activities. Environmental
Protection Publication SW-142. [Washington], U.S. Environmental
Protection Agency, Jan. 1975. 74 p.
-------
classification of solid waste, similar in concept but larger in scale
than EPA's prototype demonstration plant in St. LouisJ^ The second and
broader objective was to achieve an improved perspective on the diverse
variables that affect costs—including accounting procedures as well as
design assumptions and site-specific costing parameters—and thus to
point the direction towards more meaningful estimates in the future.
Following a brief introduction to the primary data sources, esti-
mating methods, and design assumptions, comparative results will be
presented for capital investment costs, plant operating and maintenance
costs, other special cost factors, product revenues, and a final synthesis
of net processing costs for a number of recent shredded-fuel plant
designs.
It should be emphasized that the dry-shredded-fuel system is only
one of several material and energy recovery technologies currently under
development,3 and that the cost estimates presented in this paper apply
only to the specific technology under review and only under the general
cost-accounting assumptions enumerated below. Readers are thus cautioned
to exercise care in interpreting and applying these cost estimates.
Although the estimates themselves are both technology-and time-specific,
the accounting framework and many of the procedures used to standardize
the diverse costing methods should prove applicable to all type of
systems.
CENTRAL METHODS AND DESIGN ASSUMPTIONS
What the Data Represent
The capital and operating cost estimates presented below are
derived from a comparative review of five recent preliminary engineering
designs.+ The plant designs selected are typical of improved versions
of shredded fuel plants patterned after EPA's St. Louis demonstration.
All five versions could be considered in either the medium (750 to 1,000
tons per day) or large (1,200 to 2,000 tons per day) size class by
current standards. The first commercial application for a plant of this
type is scheduled to start up in 1976.
*A1though generally considered to be in the fuel or energy recovery
category, this technology is potentially adaptable to fiber recovery for
recycling. It may also sometimes be considered as a first-stage unit in
an integrated steam or electric generating facility. In the present
study, glass and nonferrous metal recovery subsystems have been excluded
from the plant flowsheet in developing the cost and revenue estimates.
+The five design documents selected were chosen from a much larger
sample of preliminary design studies and cost proposals for plants of
tliis type. Selection was based on level of costing detail and currency
of estimates.
-------
The technical designs themselves have been partially modifed in
order to reflect a more standardized flowsheet including: hand-sorting
of paper, two-stage shredding (or milling) with one-stage air classifica-
tion to produce a marketable fuel product, and magnetic separation of
ferrous metals. Glass and aluminum recovery components have been
excluded from the present standardized flowsheet, although originally
included in some of the source designs. In addition, original cost estimates
have been "normalized" to adjust for a number of differences among the
original design studies in terms of estimating methods, accounting
formats, and site-specific cost factors.
The five original plant designs and cost estimates are attributable
to the following sources.
1. The National Center for Resource Recovery (NCRR),:in an engi-
neering feasibility study (December 1972, Ref. 4) as revised
in the Winter of 1973-74 in connection with a request for
proposals for a plant to be constructed in New Orleans, Louisiana,
(Ref. 5). (The EPA modified version is referred to below as
NCRR/EPA).
2. Midwest Research Institute (MRI), in a project performed for
the Council on Environmental Quality, completed in the Summer
of 1972 (Ref. 6), with estimates updated and revised during
the Autumn of 1973. (The modified version referred to below
as MRI/LPA).
3. The General Electric Company (GE), in a preliminary plant
design under contract to the Department of Environmental
Protection, State of Connecticut, completed in the Spring of
1973; hypothetically sited for Hartford, Connecticut, (Ref. 7).
(Modified version referred to as GE/EPA).
4 and 5. Two confidential proposals actually submitted to a city in
1974. These two designs have been merged into a composite
"Plant X" as a means of preserving the confidentiality of
proprietary information. (Referred to as X/EPA).
Before presenting the comparative cost results, further comment on
the standard plant design and the issues of normalizing costs is necessary
in order to define the scope and meaning of the estimates.
-------
Standardizing the Plant Designs
This involves either adding or subtracting building space and
equipment items. The objective is not to achieve a completely standardized
plant, but rather to standardize only the basic processing sequence and
"product lines" while preserving variations in original design conceptions
such as structural plant features, throughput and storage capacities,
number of primary process lines, and certain other special characteristics
considered important by the original designers. They still represent
different (independently produced) design conceptions for the same
general type of resource recovery facility.
In order for the cost estimates to be meaningfully comparable, it
is desirable to be able to standardize the technical assumptions or
design conditions relating to plant "capacity," annual operating schedule,
and raw waste input composition. "Capacity" turns out to be an ambiguous
variable in current design literature. Differences in specifications
regarding assumed number of hours per day and total hours per year for
plant operation typically vary among designs of the same nominal "capacity"
by a factor of two or more. For present purposes, the rated hourly
"design" throughput tonnage is accepted as given by the original source.
But it has been assumed that the plants will all operate typically on a
full two-shift (16 hours per day) processing schedule as a definition of
daily design capacity. For purposes of calculating annual fixed costs
per tons, maximum annual capacity is based on an assumed 5,000 hours at
average hourly design capacity.*
The estimates presented below also assume a standard, national
average raw waste input composition8' P-10, together with material
recovery efficiency factors as follows:
(1) 25 percent efficiency in hand-picking of old news and corrugated
paper.
(2) 90 percent efficiency in recovering organic material as fuel.
(3) 90 percent efficiency in recovering the ferrous metal fraction
as steel scrap.
*Five thousand hours is roughly equivalent to 312 days per year
(6 days per week times 52 weeks) times 16 hours per day. For a 1,000-
ton-per-day plant (62.5 tons per hour times 16 hours per day), this
implies a maximum annual capacity of 312,000 tons.
-------
Normalizing the Cost and Revenue Estimates
In addition to technical design and operating features, a very
large number of nontechnological variables and costing procedures can
also strongly influence the estimates. As far as possible, these variables
and procedures have been "normalized" to derive the present estimates.
This means that the original sources' design costs have been
recalculated on the basis of standardized price or other costing assumptions,
As noted below, a number of special cases have been identified where a
factor is both significant and can vary over a wide range. For many of
these, alternative calculations are presented to show the particular
influence of these variables at both high and low values.
The present section discusses the significance of these individual
items and their treatment in the study.
Items Affecting Capital Cost. The following items affect initial
capital investment cost and, hence, annualized capital cost per ton.*
(1) Land cost. May or may not involve initial direct financing.
May or may not be accounted for explicitly in engineering cost
estimates. Could amount to a million dollars or more.
Excluded from basic capital cost and included under other
special cost items because of extreme variations in treatment
by different sources.
(2) Site preparation. Extremely site-specific. Demolition of
existing structures could amount to several hundred thousand
dollars, and thus has been excluded from the standard capital
cost estimates.
(3) Regional construction cost differentials. Direct capital
costs typically can vary among cities between 75 percent and
115 percent of the U.S. national average. Plant costs in this
paper were adjusted to the national average base using regional
construction cost indices.
(4) Indirect construction contractor overheads and fees. May or
may not be explicitly included by different estimators. Can
by 25 percent or more of direct construction costs. In addi-
tion, architectural and engineering fees are typically 6 to 8
percent of direct costs.
*Tne conversion of an initial capital investment cost to an annual
fixed cost for accounting or debt management purposes is discussed in
•^ "1 -i +• ^ v» f s^ r* 4- -I *-i v\
a later section.
-------
(5) "Contingencies." May or may not be explicitly itemized in the
estimates. Included as a hedge against unforeseen circum-
stances in construction. Not a real cost unless some unforeseen
circumstance materializes. "Contingencies" are included in
addition to labor and equipment cost escalations per se.
May be 8 to 15 percent of total plant and equipment costs.
(6) Construction cost escalations. In effect, another type of
contingency—estimated cost increases for labor, material, and
equipment during construction period. Varies both with
length of construction period and annual percentage increase
assumed. Differences among estimating factors can cause
multi-million dollar differences in capital cost estimates.
EPA normalized estimates converted to January 1974 base
period, where necessary.
(7) Plant startup and working capital. May or may not be included.
EPA estimates normalized at four months of operating costs
capitalized with other initial investment.
Items Affecting Operating and Maintenance (O&M) Costs. O&M costs
are defined here to include only direct, pi ant-related labor, parts,
materials and supplies, and utilities. Other annual costs are included
under other special costs, discussed separately in a later section.
(1) Regional Q&M factor price differentials. Operating wage rates
can vary regionally by more than j^I5 percent of the national
average. Electric utility rates can vary by a factor of more
than 50 percent geographically; fuel prices per Btu can vary
by a factor of three or more. The O&M cost figures in Table 3
reflect such adjustments by converting to U.S. nationwide
averages.
(2) O&M cost escalations. Escalated differently by different
estimators, usually to first year of plant operation from base
date of original quote. Differences in original date, projected
startup date, and assumed rates of increases can mean a
difference of over 50 percent in total O&M cost estimates
among different sources. Standard base date of EPA-normalized
estimates is January 1974.
(3) Transport costs. Costs of transporting recovered materials
accounted for here either in estimating net selling prices or
in other special cost category. In various published sources,
they have been included under general capital and O&M accounts
or ignored altogether.
-------
Other Special Costs of Operations. Five special cost items have
been identified which, under various conditions, can each have values
ranging from zero to over $1 per ton of raw waste processed (i.e.,
$300,000 per year based on a 1,000-TPD plant operating 300 days per
year). Such wide possible variations can be either locational or
institutional in origin. "High" estimating options are indicated below.
(1) Local property taxes. Resource recovery facilities usually
have been viewed in the same category as public disposal
sites; property taxes seldom have been included in the cost
accounts. Some State and regional systems do include an
equivalent payment in lieu of taxes, based on assessed value.*
An annual charge of 4 percent on total value of property is
taken as a "high" cost factor in the comparisons below.
(2) Residual waste disposal costs. About 20 percent of weight
(perhaps 5 to 8 percent of volume) or raw waste input not sold
as product by present assumption. If disposed of as waste, a
disposal cost of $5 per ton assumed "high" for this type of
compact, shredded material (equivalent to $1 per ton of total
raw waste input). At the other extreme, glass and aluminum
content could make the material marketable.
(3) Non-pi ant^overheads. Chargeable to plant operation for off-
site services by either a private or public sector central
management agency. Could include bookkeeping, marketing,
engineering or other functional services, or general overhead.
For extreme comparisons, assume range from zero to $1 per ton.
(4) Management fees (profit). Payable to private operator of a
publicly-owned or-leased facility. Zero for a publicly-
operated facility. One dollar per ton of waste processed
would seem to be a "high" fee (exclusive of corporate overhead
expenses).
(5) Shredded product transportation costs. Depending on who pays,
could be accounted for as reduction in selling price. Treat
as separate item chargeable to shredding plant operation. For
plants located adjacent to user's boiler, transport cost can
approach zero. A "high" cost for reasonably long distances (25
miles) would be S3 per ton of output material ($2 per ton raw
wet input basis). Since this is a very large volume item,
significant annual costs are involved.
*The use of "payments in lieu of taxes" is also a means of reducing
local prejudice against the location of a regional facility in a
particular city. It is also a partial means of compensating a community
for additional implicit costs such as increased truck traffic, noise, etc
-------
Normalized Product Revenue Estimates. Given raw waste input com-
pos itTorTalTd~rl!coverYirff^^ discussed) and assuming
that product markets are available, then product revenues will be
determined by selling prices, less any relevant discounts and transport
costs.
/
Product selling prices easily constitute the greatest source of
uncertainty in the entire resource recovery picture. They exhibit the
largest variations among geographic regions at any time, and secondary
material prices historically have been subject to extreme fluctuations.
Future negotiable prices for recovered fuels and metals are also subject
to some additional uncertainties due to technical questions of product
performance (quality).*
For these reasons it was decided to develop new "high" and "low"
product revenue estimates rather than use those found in the original
source documents. The estimated revenue schedules are presented below
in Table 1. The basic assumptions and derivations of the values for the
three products are summarized in the notes accompanying that table.
The prices for both ferrous and paper are stated as values received
by the seller (processing plant) net of all transport charges. Shredded
fuel prices, however, are defined net of a power plant firing cost dis-
count (assumed at $2.50 per ton of fuel) but without deducting costs of
transporting the shredded fuel to the power plant. As previously
noted, because it can be such a large and variable element, the cost of
transporting the fuel has been singled out for special mention under the
other special costs category.
The net product selling prices are combined in Table 1 with the
product-yield assumptions to calculate revenue per ton of total raw
waste input. Thus, adding all the "high" product revenue estimates
results in a total maximum revenue of $15.85 per ton of waste processed.
This contrasts sharply with the minimum total net revenue receivable
under the low value assumptions of $3.40 per ton of waste processed.
It should be emphasized that the "high" and "low" estimates represent
neither the maximum nor the minimum conceivable under all present or
future U.S. circumstances. Rather, they simply represent the results of
a combined assessment of assumptions relating to product grading (quality)
specifications, current U.S. average fuel prices, and material prices
experienced within the past 2 years. The high estimates assume no
future increase in prices, but the low values assume that wastepaper and
steel scrap prices will not fall very much below their lowest levels of
the past 2 years. The worst case would be where no markets exist for
the shredded fuel or other product.
*A more important issue, not dealt with here, is the possible types
of long-term contractual arrangements that may be developed with user-
industries. These might eventually be able to dampen cyclical price
fluctuations and also be able to achieve higher product grade ratings
than would otrierwise be achievable in the general spot markets.
-------
TABLE 1
NET PRICES RECEIVED PER TON OF PRODUCT
AND REVENUE PER TON OF RAW WASTE PROCESSED:
"HIGH" AND "LOW" ESTIMATES (1974)*
Products
Net Prices Per
Ton of Product
Output "*
"High"
"Low"
Recovered
Product As A
Percentage of Total
Waste Input
(wet weight basis)
Net Revenue Per
Ton of Total Waste
Input
"High"
"Low"
Shredded fuel *
Paper §
Ferrous Metal '
Totals
$15.50
40.00
50.00
-
$2.50
20.00
12.00
-
67.0%
4.0%
7.7%
78.7%
$10.40
1.60
3.85
$15.85
$1.70
0.80
0.90
$3.40
*U.S. Environmental Protection Agency estimates. Office of Solid
Waste Management Programs, Resource Recovery Division.
+Prices received by seller net of transport or other discounts.
:(:Based on Btu value of shredded fuel at 10 million Btu per ton,
30 percent moisture, less $2.50 per ton estimated firing cost to user.
"High" net price based on $18.00 per ton fuel (equivalent to $1.80 per
million Btu average U.S. contract price for utility grade residual fuel
oil in Spring 1974). "Low" price based on $5.00 per ton fuel (equivalent
to coal at $0.50 per million Btu or $11.00 per ton), less $2.50 firing
cost.
§Average combined prices of old news and corrugated, F.O.B. recovery
plant, assuming buyer pays freight. "High" $40,00 price is U.S. average
in Spring, 1974. "Low" $20.00 price is U.S. average in Winter 1972-73.
Official Board Markets publication quotes.
iiAverage1 scrap steel grade better than No. 2 Bundle grade, less
$10.00 per ton freight paid by seller. Gross "high" price of $60.00,
Spring 1974 U.S. average. Gross "low" price of $22.00 is Winter 1973
U.S. average. American Metal Market publication quotes.
-------
COMPARATIVE SUMMARY OF NORMALIZED
CAPITAL INVESTMENT COST ESTIMATES
Total Capital Cost
Capital investment costs for the four case study plants in Table 2
reflect both the flowsheet revisions and the cost-estimating revisions
previously discussed. Otherwise, they reflect the differences in
design-conception of their original design teams.
The normalized capital cost estimates for the standardized plant
designs show a much closer grouping of values than do the original
capital cost figures. However, remaining differences may still seem
surprisingly large to many readers. Thus, estimated total investment
cost among the four plants varies by a factor of three, from $5.2 million
(NCRR/EPA) to $15.5 million (X/EPA) based on 1974 construction costs.
This overall difference is reduced somewhat when account is taken of
capacity differences (compare total investment cost per ton of daily
design capacity). Thus, on a per ton basis, the X/EPA plant becomes
second lowest in capital cost at $9,700 per ton of daily capacity.
Although not all differences can be explained on the basis of
available documentation, most of the $8.8 million difference between the
normalized GE and NCRR capital cost is explained by technical and architec-
tural design differences. For example, the GE design has two completely
independent process lines, considerably more material storage space (a
particularly costly item for these plants), a pit-and-crane material
feed system, and nearly twice the fully-enclosed building area (exclusive
of input and output storage) of the NCRR design.
Annualized Capital Cost
Annual capital cost is estimated on the basis of two alternative
fixed charge (capital recovery) rates: a low 10 percent rate to illustrate
the public sector finance option, and a high 25 percent rate to illustrate
annual capital cost allocation under a private industry financing option.
It should be emphasized that the 25 percent private rate includes a
built-in private profit return on the equity portion of the original
investment. The low 10 percent rate includes only interest and amortization
for an investment wholly financed by long-term, tax-free borrowing.
The apparent difference between these two alternative institutional
approaches to plant financing is quite substantial—a factor of 2.5 in
the amounts. It should be pointed out that part of this difference
represents a Federal tax subsidy for local public sector loans, i.e.,
the tax-free nature of local government bonds.
10
-------
TABLE 2
NORMALIZED CAPITAL INVESTMENT COST
ESTIMATES FOR FOUR DRY-SHREDDED-FUEL
PROCESSING PLANT DESIGNS * +
Plant Capacity and Investment NCRR/EPA
Cost Measures
MRI/EPA
GE/EPA
X/EPA
Plant Capacity Factors:
Number of Process Lines One One Two Two
Design Tons Per Hour 62.5 62.5 62.5 100
Design Tons per Day (16 Hours) 1000 1000 1000 1600
Design Tons Per Year (5000 Hours) 312,500 312,500 312,500 500,000
Normalized Capital Investment:
Total:
1974 (Thousands) $5,200 $11,600 $14,000 $15,500
1976 (Thousands)* 5,980 13,340 16,100 17,830
Total Per Ton Daily Capacity
1974 (Thousands) 5.2 11.6 14.0 9.7
1976 (Thousands) 5.98 13.34 16.1 11.14
Annualized Capital Cost:
@ 10% per year:
1974 (Thousands) 520 1,160 1,400 1,550
1976 (Thousands)* 598 1,334 1,610 1,785
@ 25% per year:
1974 (Thousands) 1,300 2,900 3,500 3,875
1976 (Thousands)* 1,495 3,335 4,025 4,460
Capital Cost Per Ton Raw
Waste Processed ($ 1974 Base)
(P 10% Capital Charge, and:
(1) 90% capacity utilization
(2) 75% capacity utilization
(3) 60% capacity utilization
@ 25% Capital Charge, and:
(1) 90% capacity utilization
(2) 75% capacity utilization
(3) 60% capacity utilization
$ 1.85
2.20
2.75
4.60
5.55
6.85
* 4.15
4.95
6.10
10.35
12.35
15.25
$ 5.00
5.95
7.35
12.50
14.90
18.40
$3.45
4.15
5.15
8.60
10.35
12.90
*0ffice of Solid Waste Management Programs, Resource Recovery
Division. Based on original plant design cost estimates by the National
Center for Resource Recovery (NCRR), Midwest Research Institute (MRI),
the General Electric Co. (GE), and other proprietary sources ("X").
+A11 plants utilize two-stage shredding and air classification,
with magnetic separation of ferrous material and hand picking of paper.
Glass and nonferrous recovery options not included. Shredded fuel
transport facilities and land costs not included.
$1976 values escalated at 1.15 x 1974 values to account for inflation
to midpoint of construction period.
n
-------
Capital Cost Per Ton
Capital cost per ton is shown in Table 2 on the basis of the two
alternative fixed charge rates and three alternative capacity-utilization
rates. The latter are based on a somewhat arbitrary maximum design
capacity utilization of 5,000 hours per year. Ninety percent capacity
utilization probably represents a high design target from a practical
standpoint. The various lower rates can reflect a combination of an
intentionally restricted operating schedule (fewer hours per day or days
per week), additional equipment downtime for unscheduled repairs, or
restricted throughput rates due to low raw waste deliveries or output
market bottlenecks.
Other things being equal, unit capital costs will be about 20 percent
higher at a 75 percent capacity rate than at a 90 percent rate, and
about 25 percent higher still if the plant utilization rate falls to
60 percent. Overall, the difference between achieving only a 60 percent
rate as opposed to a 90 percent rate is a capital cost per ton penalty
of 50 percent. As shown in Table 2, this penalty varies in absolute
dollar terms from a low of just under $1 per ton (NCRR/EPA at 10 percent
capital charge) up to a high of almost $6 per ton for the high capital
cost GE/EPA plant (under the 25 percent capital charge rate). At the 10
10 percent charge rate, this factor alone accounts for differences of up
to $2 or more per ton for the MRI and GE designs. Even the outwardly
small differences of 75 vs. 90 percent or 60 vs. 75 percent capacity
utilization result in cost differences of $0.35 to $1.60 per ton for the
plants in our sample group. At the higher 25 percent fixed charge rate,
the effect of capital utilization rates is magnified 2.5 times.
COMPARATIVE SUMMARY OF NORMALIZED
O&M COST ESTIMATES
Table 3 provides a comparison of the O&M cost estimates for the
four preliminary designs, adjusted to account for certain design
standardizations and revised to reflect 1974 base-year national average
labor and utility cost factors.*
*It should be recalled that O&M costs do not include an item for
capital charges (or "capital recovery"). Nor do they, at this point,
reflect any adjustments either for dump fees charged to those delivering
solid wastes or revenues received from product sales. In other words,
they represent only the on-site labor, material, and utility costs of
the processing facility.
12
-------
TABLE 3
NORMALIZED OPERATING AND MAINTENANCE COST
ESTIMATES FOR FOUR DRY-SHREDDED-FUEL PROCESSING
PLANT DESIGNS*
Plant Capacity and O&M
Cost Measures
NCRR/EPA
MRI/EPA
GE/EPA
"A"
Plant
X
Plant Capacity Factors:
Number of Process Lines One
Design Tons Per Hour 62.5
Design Tons per Day (16 Hours) 1000
Design Tons Per Year (5000 Hours) 312,500
Total Annual O&M Costs:
1974 (Thousands)
@ 90% Annual Capacity Utilization $1,288
@ 75% Annual Capacity Utilization 1,128
-------
Two features of the resulting normalized O&M cost estimates are
worth special attention. The first is the relatively close grouping of
the estimates for the different plants. Thus, for a given base year,
say 1974, and a given relative operating level (say the 90 percent
capacity rate), the unit cost estimates differ by not more than about $1
per ton (20 percent). This represents a surprisingly close agreement
among the different sources, especially considering that there is so
little real operating experience upon which to base estimates.
The second general conclusion is that if the estimates for the
several plant capacity utilization rates are accurate, the unit operating
costs are moderately responsive to changes in operating levels. Thus,
the O&M cost variation for a given plant over its operating range between
60 and 90 percent of its rated capacity was estimated at about $1 per
ton (in 1974 dollars) for all four of the plants. However, the engineering
data on which the O&M cost penalties for under-capacity utilization are
based are quite sketchy. There are no published estimates or analysis
of this relationship, but it warrants more attention.
SUMMARY OF TOTAL AND NET COST ESTIMATES
The final synthesis of cost and revenue estimates is presented in
two steps. The first step, summarized in Table 4, combines the three
categories of costs (capital, O&M, and other special costs) into a range
of total cost estimates for each of the four designs in our sample. The
second step combines the total cost and revenue estimates into a set of
net cost (or net revenue) results, as illustrated in Table 5.
Total Cost Estimates
In the first part of Table 4, capital costs from Table 2 are added
to basic O&M processing costs from Table 3. The resulting "total
processing costs" are unique for each of the four preliminary plant
designs. Basic processing costs are estimated to range from $6.45 per
ton for NCRR/EPA to $10.55 for GE/EPA at the low (10 percent) capital
charge and the high (90 percent) utilization rate. At the other extreme
(high capital charge and low utilization rate), these basic costs are
90 to 150 percent higher, depending on design.
Total process cost differences among the four plants represent
differences within the engineering design community as to the capital
and operating resource requirements to process mixed waste at the
indicated scales. These are differences remaining after our recalculations
to standardize design and costing parameters. Considering the state of
technological development, the differences in process cost estimates
among the four designs are less than might have been expected. In fact,
the differences among plants due to different designers are less than
the differences for any given plant due to alternative capital charge
and operating rate assumptions.
14
-------
cu
to
ro
CO
G£ *
r> co 4J
o z tn
U. CD O
1-1 O
o; to
O Ul KC
Lu O r-^
cn
to 1— i—
LU z
1— *f "
< _i 4->
S Q- =
1— 1 CL
H- CD C
i/i z: i— i
UJ •— >
to cu
1 — CO •*-*
CO UJ 10
o c_> 10
O O 3
"S
O 0- 2
UJ IO
h*4 .J OL.
1— < UJ
— 1 ZJ f-
41 U- O
ol a c
O LU O
Z O 1 —
a
LL. LU S-
o a; cu
1C Q.
>- CO
o_. i to
< >- s-
2 o: (0
S o •—
to o
Q
<
4-
S*
O
VO
4-
&*
€O
01
4-
3-
c cu
° 01
+•> ro
CL-C
00
CU r—
O (O
C 4J
H3 •!—
c: a
•c- ro -4->
U. CO 10
I.- -S
O 03 +J
+J 3 .— 10
O C 1C O
CU C 4-><->
to ec -r-
CLS:
oo vohi-
LO LO|O LO
ro r^ r— o
o <^-!LT> vo
LO LO O
CM VO Ol
co in ro
^~
in O|LO in
OO LO CO VO
VO LO|CM LO
*~~
LO LO O
«a- r— vo
VO LO i—
O O O LO
VO VO CM i—
•* «-f Ol LO
C
0'—
••- 0)
+-* Ol
4-> CL S- <-)
to CU o ro to CU
O .Q C CU O (S S
01 "+!o £ = •- to 'w^
totoy; raio tototo
CUOTTS- C-l-> CUO+->
OD-«CU -r--r- UO.W
p°o_ U-Q+J o o
S-S-^-1 rotos-s_cj
Q-CU 4-> S-O O Q-CU
XI r— 10 O O .C r-
r- -Uroo -P.— ^j^^jnj
lOO-i-O Oro i— lo ro O T-
+JCJ CU3roO+->O
Oi — CUi — tOC 4->OOi — CU
•MlOCLra CT- 4->rOCL
XJ +-> tO +J CU
CM
LO
CM
O
o
CO
r^ '
LO|
rr>
«*
i— 1
1
c
o
1—
i.
cu
Q-
+J
CO
S
r—
ro
4->
O^S I0033O O 4J IOOO3O ~*
••- O CJOIOI— 1— (O&S CJOtOH-
JD
3 C&
O- •
> LO
•i- CM
S-
CL. c°j
+J
in
>> cu
•i-J s-
i- QJ
CU +J
C Q. C -(->
cn o ••- re
•r- L.
to CU Q. i. 4-> CU
CU -C ro S- 4->
•o +J -— . cu o to
.— >, D. ro
4_> M — - m 2
C ^— • S- C
ro >-> •- CU ro 2
"CLZ: c: *-> s_
— * -r- r~.
>— -a o r— li-
ra CU 3 • CU O
C +•> •— O 3
•i-3 4- C
O14J C +J O
•r- -i- -I- IO T3 4->
S- 4-> CU
C in O T3 CU
C i— i +J -i- CU CL
o in i— s-
.c: o i— x: o
T3 O (J -r- to O
CU S- E
CO ro CU • — -I —
ro cu r— vo *r vt
CQ to .Q .~_-
CU •!- 1— f-
a; to*q- •• o
to ^^
c 4-1 o i- c: to
o to CL o o cu
•r- CU 4-> 01
CO S S- -^ S-
•i- -O CU S- 4-> ro
> -i- J= O 3 -C
•i- s: 4-> 2 a. o
o o c
" CU -i- T3
>,'-> Ol+J ro
L- QiC C *r~ 2 CU
cu aC -i- to ro -c
> o • 2 s- s.
0 Z-— Oi— CU
0- — = i— ro S- >
CU X r— 3 CU O
cf. •>,- o to CL
S. "4-3 +•>
CU O) C O C
O > to CU 3 O ro
s_ o ai .c • i—
3 O U 4-> T3 i— D.
o cu s- c: to- i
to a: 3 i|_ ,0- — • c
cu o o o
Q: CU I/I T3 CU C
(j a> c 4->
*• J- >» > ro tn ^~-
to 3 S- •!- r— ro in
E 0 ra <*_ 2^
ro to 4-> --^
i- CU CU • i— CM C T3
010: "- c r— -~^ o c
£1- O ro +•> JO
s_ a..,-
Q-OO4->I-4J1."-
it- 1. ro o C CU^ — •
4-> CL NJ It- (U CL C
C S- -i- S O
f4J CU T- CU to O
C -C •(-> 3
O1CU4J 3i— >LO 3
ro O O IO C 0=* Q.
C >)>••- C
IO i— -O 4J 4-> -i-
.E ro C ••- -a r— re
C ro O CU ro 2
CU O ra 4-> +J .— ro
4->*i— "C-LroOroS-
tO 4-) ^'^ IO £ 4-' IO
ro ro LU O T- O S-
3 Z CD 4->4- 0. cu
4J to o to CL
-a cu c cu ••-
••- J= -10 4-> T3 O
•— 4J O •— = C O
O O CL.C CU CU •
00 >, Ol O 4-> CM
^.1 O <+- .r- J_ in **
OCt)-Or— O
U ro CU CU O • ro 4->
•i- E r— O T.)- 3
4- -i- UJ S_ E "OI.-O
4- +J CU .3 +J -i- CU CU
Ot/li — O_l/>rOtOCLi/)
* CU O) o U
to c: x • — ^ • o
O cu ro ro ro I.
O CD 4J • 0^- Q.
15
-------
TABLE 5
SUMMARY OF ALTERNATIVE NET REVENUE (COST) CALCULATIONS FOR FOUR PRELIMINARY PLANT DESIGNS
AT TWO ALTERNATIVE CAPACITY UTILIZATION RATES
(Dollars Per Ton of Raw Waste Input, 1974 Cost Basis}*
NCRR/EPA
MRI/EPA
GE/EPA
60%
"X'VEPA
High Revenue Cases:
Case 1: High Revenue Estimate
with Process Cost Only.
Total Product Revenue $15.85 $15.85 $15.85 $15.85 $15.85 $15.85
Less. Total Process Cost+ 6.45 8.25 8.90 11.80 10.55 14.00
Less: Min. Other Special
Costs
Equals: Net Revenue
Case 2: High Revenue Estimate
with Maximum Other Special Costs.
Total Product Revenue $15.85 $15.85 $15.85 $15.85 $15.85 $15.85
Less: Total Process Cost+ 6.4b 8.25 8.90 11.80 10.55 14.00
Less: Max. Other Special
Costs 5.15 5.65 6.05 7.00
Equals: Net Revenue (Cost)
Low Revenue Cases:
Case 3: Low Revenue Estimate
with Process Cost Only.
Total Product Revenue $ 3.40 $ 3.40 $ 3.40 $ 3.40 $ 3.40 $ 3.40
Less: Total Process Cost+ 6.45 8.25 8.90 11.80 10.55 14.00
Less: Min. Other Special
Costs
Equals: Net Revenue (Cost)
$15.85 $15.85
8.35 10.95
$15.85
8.35
5.65
$15.85
10.95
6.40
$ 3.40 $ 3.40
8.35 10.95
($ 4.95) ($ 7.55)
Case 4: Low Revenue Estimate
with Maximum Other Special Costs.
Total Product Revenue
Less: Total Process Cost+
Less: Max. Other Special
Costs
Equals: Net Revenue (Cost)
3.40
6.45
$ 3.40
8.25
$ 3.40 $ 3.40
8.90 11.80
$ 3.40
10.55
$ 3.40
14.00
$ 3.40 $ 3.40
8.35 10.95
*0ffice of Solid Waste Management Programs, Resource Recovery Division. Based on original plant design
cost estimates by the National Center for Resource Recovery (NCRR), Midwest Research Institute (MRI), the
General Electric Company (GE), and other proprietary sources ("X").
+Sum of capital cost and O&M cost from Table 4. Capital cost based on 10.0 percent annual fixed charge.
rate (capital recovery).
16
-------
As previously discussed, the other special cost items may or may
not be relevant under particular locational and institutional circumstances,
Thus, each of these cost items may have zero values for particular
cases, or they each may involve substantial additional annual and per
ton expense to the recovery operation. The values included in Table 4
are our EPA "high" cost estimates. They do not necessarily reflect
either the particular values or, in some cases, even the same categories
of costs estimated in the original design source documents. Rather,
they have been applied to all the designs in our sample as an added
means of normalizing the estimates for comparative purposes.
Thus, the other special cost elements, taken as a group, can sum up
to any value from zero to some significantly higher cost. The maximum
value for our comparative cases varies between $5.15 and $7.50 per ton,
depending on plant capital cost (a variable in the property tax cost
function) and level of capacity utilization.
In the very special case where other special costs are all zero,
total processing cost is the only cost to be balanced against product
revenues to determine net cost or revenue from plant operation.
Net Revenue or Cost Results
The final step in the cost-estimating procedure is to subtract
total cost from product revenues to yield net revenue (profit) or cost
results.* Table 5 presents four sets of net cost calculations for each
of the four case study designs to show the various combinations of:
high revenue with low cost; high revenue with high cost; low revenue
with low cost; and low revenue with high cost.
The first two net revenue calculations for each plant represent the
low and the high cost possibilities as developed in Table 4 in conjunction
with the "high" ($15.85 per ton) total revenue estimate from Table 1.
The net revenue line for Case 1 indicates positive net revenues for all
plants. Thus, so long as "high" revenues can be combined with costs
that do not exceed standard process cost by substantial amounts, all
four case study plants appear profitable at the current estimated
values under public sector financing. Even when a maximum other special
cost sum is charged (Case 2), NCRR/EPA remains profitable at the 60
percent capacity utilization rate, and both MRI and Plant X continue to
show net revenue at high utilization rates.
*0ther things being equal, in situations where cost exceeds product
revenues, the net cost values may be considered equal to the dump fee
required for the facility to break even. Alternatively, one may wish to
compare these net cost values against the community's alternative
opportunity costs of conventional disposal or other resource recovery options
17
-------
For the low revenue ($3.40 per ton) Cases 3 and 4, net positive
revenue disappears, even where low costs are involved. Results show net
costs of about $3.00 to $7.00 per ton at 90 percent utilization rates
and $5.00 to $11.00 at low capacity rates for Case 3. It is noteworthy
that the net costs in this line are still generally competitive with
landfill costs in many, if not most, highly urbanized areas.
The final "bottom line" (Case 4) represents the worst situation
shown with respect to resource recovery--i.e., low revenue combined with
highest possible cost for the plant cases presented. Even the results
for these worst-case resource recovery alternatives are encouraging
because net cost estimates in all cases remain competitive with
conventional incineration.
A number of caveats must be made. The first is that the results in
Table 5 all assume the low (public sector) 10 percent capital recovery
rate. Costs increase under a strict private-enterprise rate of return
formulation. However, a privately-financed facility, if well managed
and strategically located, could be profitable under some realistic
locational and market circumstances. Another point that must be kept in
mind is that all the basic cost estimates are themselves subject to
substantial possibilities for error. No such plant has yet been constructed
or operated, and all costs are based on preliminary design estimates
rather than final detail design figures. Further, a serious effort has
been made to present costs on a national average basis, and many of our
urban areas will have costs at least 10 to 15 percent higher than these
estimates on the basis of location alone.
Finally, it should be noted that the present analysis does not
evaluate the question of "economies of scale" for plants of different
design capacities. Generally, one would expect that, other things being
equal, plants smaller than those in the study sample would show higher
capital and operating costs per ton than the estimates presented here,
and conversely, larger plants might result in somewhat lower unit costs.
However, an analysis of economies of scale is beyond the scope of this
study, and there has been no definitive quantitative work on this
subject to date.
SUMMARY AND CONCLUSIONS
The Environmental Protection Agency has analyzed a number of
engineering design conceptions for the next generation of shredded fuel
recovery plants based on the St. Louis prototype. Existing cost estimates
prepared by engineering consultant and system development companies are
not directly comparable with one another because of differences in esti-
mating methods, accounting formats, and location-specific costing factors.
Therefore, five recent preliminary design cost studies were normalized
to produce comparable cost estimates representative of the degree of
consensus within the engineering community.
18
-------
The results indicate that differences in cost estimates among
design conceptions and engineering firms are still quite significant,
even after adjustments for location, time, and other nonstandard elements
of costing procedure. However, the differences are no greater than
might be expected given the present state of technological development
and lack of operating commercial prototypes. Indeed, differences in
basic capital and operating costs attributable to different technical
engineering conceptions are in many respects of less consequence than
the differences introduced by the use of alternative costing methods and
location-specific cost factors.
Analysis of normalized cost estimates and alternative product
selling price projections indicates that potential net cost projections
will fall in a very broad range from positive to negative. The results
suggest that there could be some favorable cases where operation of this
type of processing plant will yield a profit from sales of product,
exclusive of dump fees. Intermediate cases-~i.e., those which combine
either high revenue with high cost or low revenue with low cost—generally
appear competitive with current or projected landfill costs in many, if
not most, U.S. cities. All low cost (public sector) financing options
were at least competitive with conventional municipal incineration, even
for the highest cost case study plant.
For a project planning and evaluation standpoint, three conclusions
of the analysis bear special emphasis:
1 • The relative importance of total revenue and the very large
absolute differences between high and low estimates. The most
significant aspects of uncertainty relate to the largest
volume output, namely, the potential market value of the
shredded fuel. Differences between "high" and "low" shredded
fuel selling price estimates account for most of the difference
between a $16 and a $3 total product revenue per ton of raw
waste processed. This difference dwarfs almost all other
variable elements of the net cost and revenue estimates.
2. The significance of maintaining high capacity utilization
rates. This is evident in the comparisons for individual
plants where differences in net cost of $2 to over $4 per ton
consistently result for estimates at the 90 percent vs. 60
percent capacity utilization rates. The high cost of failure
to maintain high capacity utilization levels underlines the
importance of sound planning and high quality management.
3. The cumulative importance of other special cost elements. If
costs are divided into three categories as in Table 4, it
comes as something of a surprise that other costs can be
larger in total than either the standard capital cost or the
direct O&M processing cost categories. The potential cumulative
effect of these items on the overall net cost picture suggests
that they are worthy of considerable attention by planners and
designers from a cost minimization standpoint.
19
-------
REFERENCES
1. Smith, F.A. An evaluation of the cost of recovering dry-shredded
fuel and material resources from mixed community solid waste.
Washington, U.S. Environmental Protection Agency, Office of Solid
Waste Management Programs, Resource Recovery Division, Aug. 20, 1974.
various pagings. (Unpublished report.)
2. Lowe, R.A. Energy recovery from waste; solid waste and supplementary
fuel in power plant boilers. Environmental Protection Publication
SW-36d.ii. Washington, U.S. Government Printing Office, 1973. 24 p.
3. Levy, S.J. Markets and technology for recovering energy from solid
waste. Environmental Protection Publication SW-130. Washington,
U.S. Environmental Protection Agency, 1974. 31 p.
4. Materials recovery system; engineering feasibility study. Washington,
National Center for Resource Recovery, Inc., Dec. 1972. various
pagings.
5. Cost analysis for the New Orleans resource recovery and disposal
program. Washington, National Center for Resource Recovery, Inc.,
1974. 108 p.
6. Franklin, W.E., et al. Resource recovery processes for mixed
municipal solid wastes; part I—technical review and economic
analysis. Environmental Protection Publication SW-101. [Cincinnati],
U.S. Environmental Protection Agency, 1973. 67 p.
7. Godfrey, D.E., et al. [General Electric Company]. Preliminary design
of a solid waste separation plant; final report. Hartford, Conn.,
State of Connecticut Department of Environmental Protection, July
1973. 208 p.
8. Smith, F.A. Comparative estimates of post-consumer solid waste.
Environmental Protection Publication SW-148. [Washington], U.S.
Environmental Protection Agency, May 1975. 18 p.
* US GOVERNMENT PRINTING OFFICE 1975- 632-820/36 20
------- |