EPA 910/9-82-0896
United States
Environmental Protection
Agency
Region 10
1200 Sixth Avenue
Seattle WA 98101
Air & Waste Management Division February 1984
?/EPA Residential Wood
Combustion Study
TaskS
Wood Fuel Use Projection
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 3
WOOD FUEL USE PROJECTION
-------
RESIDENTIAL WOOD COMBUSTION STUDY
TASK 3
WOOD FUEL USE PROJECTION
-FINAL REPORT-
PREPARED BY:
William T. Greene,President (Task 3 Manager)
Solutions for Energy and Environment, Inc. (SE2, Inc.)
1615 NW 23rd Avenue
Portland, Oregon 97210
AND
Robert L. Gay Ph.D., Consultant
4423 SW Hamilton Terrace
Portland, Oregon 97201
PREPARED FOR:
DEL GREEN ASSOCIATES, INC.
ENVIRONMENTAL TECHNOLOGY DIVISION
1535 N. Pacific Highway
Woodburn, Oregon 97071
Under Contract No. 68-02-3566 FROM:
U.S. ENVIRONMENTAL PROTECTION AGENCY
Region X
1200 Sixth Avenue
Seattle, Washington 98101
Task Manager
Wayne Grotheer
December, 1982
-------
THIS REPORT CONSISTS OF SEVERAL DIFFERENT PARTS.
THEY ARE LISTED BELOW FOR YOUR CONVENIENCE.
EPA 910/9-82-089a Residential Wood Combustion Study
Task 1 - Ambient Air Quality Impact
Analysis
EPA 910/9-82-089b Task 1 - Appendices
EPA 910/9-82-089c Task 2A - Current & Projected Air Quality
Impacts
EPA 910/9-82-089d Task 2B - Household Information Survey
EPA 910/9-82-089e Task 3 - Wood Fuel Use Projection
EPA 910/9-82-089f Task 4 - Technical Analysis of Wood Stoves
EPA 910/9-82-089g Task 5 - Emissions Testing of-Wood Stoves
Volumes 1 & 2
EPA 910/9-82-089h Task 5 - Emissions Testing of Wood Stoves
Volumes 3 & 4 (Appendices)
EPA 910/9-82-089i Task 6 - Control Strategy Analysis
EPA 910/9-82-089J Task 7 - Indoor Air Quality
-------
DISCLAIMER
This report has been reviewed by Region 10, U. S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE OF CONTENTS
page
Executive Summary viii
List of Tables i'ii
List of Figures v
I. Introduction 1
A. Purpose 1
B. Summary of Technical Approach 1
C. Overview of Report 4
II. Analysis of Short-Term Wood Fuel CTse Trends 6
A.' Short-term Trend Parameters Considered 7
Quantifying Firewood Use 7
Indirect Indicators of Wood Fuel Use ' 19
Direct Indicators of Effects of Wood Fuel Use 21
B. Findings and Conclusions 26
C. Best Estimate of Short-term Wood Fuel Use Trends 29
III. Analysis of Long-Term Wood Fuel Use Trends 31
A. Methods for Long-term Wood Fuel Use Projections 32
B. Description of Selected Analytical Approach 37
Description of Marshall's Model 37
Addition of Fireplace Usage Sector 38
Model Calibration 41
Model Sensitivity Analysis . .48
C. Findings and Conclusions 52
Portland Metropolitan Area 54
City of Seattle 56
City of Spokane 58
IV. References 60
APPENDIX A: Model Documentation
APPENDIX B: Major Factors Not Included in Marshall's Model
11
-------
LIST OF FIGURES
page
Figure 1 - Recent Trends in Firewood Volumes Removed 16
under Wood Cutting Permits from National
Forests near Portland, Seattle, & Spokane
Figure 2 - Twenty Year Trend in Number of Wood-burning 20
Stove Type Devices Shipped in the USA
Figure 3 - Recent Trends in Normalized Nephelometer 28
(Bscat) Data for Several Portland and
Seattle Sites
Figure 4 - Basic Assumptions in Marshall's Wood Use 39
Projection Model
Figure 5 - Portland Metropolitan Area Total Wood Use, 43
Fireplace Wood Usage, and Stove Wood Usage
Figure 6 - Portland Metropolitan Area Stove Sales and 44
Stove Wood Usage from 1970 to 1982
Figure 7 - Base Case 1970-2000 Projection for the 46
Portland Metropolitan Area for Stove and
Furnace Wood Usage, Fireplace Wood Usage,
and Total- Wood Usage
Figure 8 - New England 1970-2000 Projection for Wood 49
Usage in Stoves and Furnaces
Figure 9 - Portland Metropolitan Area 1970-2000 Wood 51
Use Projections with Constant Wood Prices
from 1980- 2000
Figure 10 - Portland Metropolitan Area 1970-2000 Wood 53
Use Projections with 5% Per Year Real Wood
Price Escalation from 1980-2000
Figure 11 - Wood Installations vs. Payback Period A-18
Figure 12 - Model Estimated Stove Wood Usage in Seattle A-32
City, 1970 - 1982
Figure 13 - Model Estimated Wood Usage by All Appliances A-34
in the Seattle Area, 1970 - 1982
-------
LIST OF FIGURES (Cont.
page
Figure 14 - Base Case Model Projected Wood Usage in A-35
Seattle City, 1970 - 2000
Figure 15 - Model Estimated Wood Usage by All Appliances A-45
in the City of Spokane, 1970 - 1982
Figure 16 - Base Case Model Projected Wood Usage in the A-46
City of Spokane, 1970 - 2000
Figure 17 - Oregon Unused Wood Residue Quantities B-2
Figure 18 - Washington Unused Wood Residue Quantities 8-3
IV
-------
LIST OF TABLES
page
Table 1 - Best Estimate Projections of Residential Wood X
Fuel Use for Portland, Seattle, and Spokane
and Corresponding Participate Emissions
Table 2 - Estimated Wood Cutting Permits and Firewood 11
Volumes Removed from Selected Federal, State
and Private Sources near Portland, Seattle,
and Spokane
Table 3 - Summary of Short-Term Trend Parameters Quantified 27
Table 4 - Projected Range of Year 2000 Jood Usage, by 47
Appliance, for the Portland Metropolitan Area
Table 5 - Best Estimate Projection of Residential Wood 54
Usage for Portland for 1985 - 2000
Table 6 - Change in Portland Emissions from Residential 55
Wood Burning if DEQ Emission Factors Remain
Constant
Table 7 - Projected Range of Year 2000 Wood Usage, by 56
Appliance, for the City of Seattle
Table 8 - Best Estimate Projection of Residential Wood 56
Usage for the City of Seattle from 1985-2000
Table 9 - Change in Seattle City Emissions, from Residential 57
Wood Burning if DEQ Emission Factors Remain
Constant
Table 10 - Projected Range of Year 2000 Wood Usage, by 58
Appliance, for the City of Spokane
Table 11 - Best Estimate Projection of Residential Wood 58
Usage for the City of Spokane from 1985-2000
Table 12 - Change in Spokane City Emissions from Residential 59
Wood Burning if DEQ Emission Factors Remain
Constant
Table 13 - Portland Metropolitan Area Consumer Price Index, A-3
1970 - 1980
-------
LIST OF TABLES (Cont.
page
Table 14 - Historical Portland Area Oil Prices in Actual A-4
and 1980 Dollars
Table 15 - Historical Portland Area Gas Prices in Actual A-5
and 1980 Dollars
Table 16 - Portland Area Actual Electricity Prices, 1970-1980 A-5
Table 17 - Portland Metropolitan Area Weighted Average A-6
Electricity Costs in Actual and 1980 Dollars
Table 18 - Wood Prices 1970-1980 in Actual and 1980 Dollars A-7
Table 19 - Portland Area Shares of Household Heating Via A-7
Conventional Fuels, 1970 - 1980
Table 20 - Portland Area Projected Shares of Household A-8
Heating Via Conventional Fuels, 1985 - 2000
Table 21 - Projected Portland Area Future Residential Oil A-9
and Gas Prices
Table 22 - Future Portland Area Electricity Price Projections A-9
Table 23 - Portland Area Wood Price Projections 1980 - 2000 A-10
Table 24 - Efficiency of Heating with Conventional Fuels A-ll
Table 25 - Market Penetration Table ' ' A-12
Table 26 - Variables Modified in Portland Model Runs A-12
Table 27 - Portland Area Normal Wood Installation Cost A-13
Versus Capacity Fraction Installed
Table 28 - New England Average Fuel Cost Savings Versus A-14
Capacity Fraction Installed
Table 29 - Portland Area Average Fuel Cost Savings Versus A-14
Capacity Fraction Installed
Table 30 - Relationship Between Capacity Fraction Installed A-15
and the Cost of Inconvenience of Wood Due to
Capacity Utilized
VI
-------
LIST OF TABLES (Cont.)
page
Table 31 - The Effect of Self Cut Wood on Price Versus A-16
Wood Heating Penetration
Table 32 - Wood Installations Versus Payback Period A-19
Table 33 - Seattle Area Population and Household Projections A-22
Table 34 - Seattle Metropolitan Area Consumer Price Index, A-23
1980 - 1980
Table 35 - Historical Oil and Gas Prices in the Seattle Area A-23
Table 36 - Historical Electricity Prices in the Seattle Area A-24
Table 37 - Historical Seattle Area Wood Prices A-24
Table 38 - Historical and Projected Fuel Use Split in the " A-25
Seattle Area
Table 39 - Projected Real Prices for Oil and Gas in the A-26
Seattle Area
Table 40 - BPA Electricity Price Growth Rates and Resultant A-27
Seattle Area Projections
Table 41 - Projected U.S. Growth in Demand for Various Timber B-5
Products, 1976 - 2000
vn
-------
EXECUTIVE SUMMARY
This report analyzes trends in wood fuel use within the city limits
of Seattle and Spokane (Washington) and in the greater Portland (Oregon)
metropolitan area. Short-term (through 1985) and long-term (through 2000)
trend projections are presented.
The following trend parameters were used in the short-term projections:
1. Volumes of firewood (cords/year) removed from the nearest national
forest(s); and
2. Average heating season nephelometer light scattering coefficient
(Bscat), normalized for temperature and meteorological effects for
two sites in the Portland area and one site in Seattle.
Other short-term trend parameters were considered but not used because of
inadequate data (not enough years of data for trend analysis, not directly
linked to levels of woodburning, missing data). These include the following:
1. Various surveys of household wood use
2. Sales of woodburning appliances
3. Census data
4. Air quality data including coefficient of haze, soiling index,
benezene soluble particulate extractions, organic and total carbon
measurements on particulate (HiVol) samples, chemical mass balance
(CMB), and carbon-14 measurements on particulate samples.
For all three areas, the best estimate of recent (1978-1981) short-term
trends was a 6-8% annual average increase in wood fuel use. At this annual
rate of increase, wood fuel use would grow by 34-47% from 1980 to 1985.
For long-term projections of wood fuel use, a state-of-the-art model
was adapted and applied to simulate wood fuel use during 1970-2000. This
model calibrated well against limited available data on actual wood use.
VI 1 1
-------
Historical wood usage between 1970 and 1980 is well predicted by the model.
The model's predictions for 1981-83 are consistent with the short-term trend
analysis and projections in Section II of this report and provided reasonable
results in limited sensitivity tests.
The model's input data and assumptions take into account estimated costs
of wood and alternative fuels, population and household growth projections,
household heating requirements, area mix of home heating fuels, and other
factors influencing residential wood fuel use. A major shortcoming of the
model is that the effect of increasing competing demands on wood supplies
is not taken into account.
The model projects increasing or decreasing wood fuel use based
primarily upon the magnitude of potential fuel cost savings from heating
with wood versus other fuels. This drives the simulated installation of new
wood burning devices and also determines the assumed rate of capacity utiliza-
tion (percent of total possible usage) of wood heating systems.
A major modification of the model was made to include fireplace wood
use. The original model .only projected wood fuel use i-n stoves and furnaces.
However, survey data for Portland indicated that over half of the firewood
used in 1978-79 was burned in fireplaces. Thus, the modified model predicts
wood fuel use (cords/year) for stoves, furnaces, and fireplaces, and their
total. The model is fully documented in Appendix A (and Reference 15),
including description of input data and assumptions.
Table 1 summarizes the model's best estimates for long-term wood fuel
use for Portland, Seattle and Spokane for 1980-2000 and corresponding
particulate emissions. Projections were made for the city limit area of
IX
-------
TABLE 1
Best Estimate Projections of Residential Wood
Fuel Use (103 cords/year) for Portland, Seattle, Spokane
(1980-2000) and Corresponding Participate Emissions (10-3 tons TSP/yr)
Y Number of Stove/Furnace
Households Wood Usage
PORTLAND METROPOLITAN AREA
1980
1985
1990
1995
2000
CITY
1980
1985
1990
1995
2000
CITY
1980
1985
1990
1995
2000
471,850
537,800
603,750
669,700
735,650
OF SEATTLE
220,000
246,180
269,720
294,360
323,180
OF SPOKANE
. 70,920
77,940
83,960
90,910
98,860
150
240
240
300
340
45
85
85
90
85
28
42
45
51
54
Firepl ace
Wood Usage
190
190
170
150
140
110
100
100
100
100
93
84
81
78
75
Total
Wood Usage
340
430
410
450
480
155
185
185
190
185
121
126
126
129
129
Total Participate
Emiss-ions
9.3
12.8
12.5
14.5
15.9
3.7
5.1
5.1
5.3
5.1
2.7
3.1
3.2
3.4
3.4
-------
Seattle and Spokane, and the greater Portland metropolitan area because
input data were more readily available.
For the Portland metropolitan area, total wood usage is projected to
increase at an annual rate of 6.5% between 1980 and 1983, 0% between 1984
IT-
i'- and 1988, and 1.6% per year between 1990 and the year 2000. This represents
L'J .
a 41% increase in total wood fuel use between 1980 and 2000 resulting from a
I
{... 127% increase in wood usage in stoves and furnaces and a 26% decrease in
; fireplace wood usage during 1980-2000. Because stoves emit more than twice
I
as much particulates per ton of wood buraed than fireplaces, the 41% increase
j in total wood usage represents about a 71% increase in particulate emissions
i
by the year 2000 with more than half occurring by 1985. The increase in
r
[_ particulate emissions may well be less than projected if more efficient
stoves become available and public education campaigns on proper burning
methods are successful.
For Seattle residences, total wood fuel use is projected to increase
by 26% between 1980 and 1985. Wood fuel use is projected to peak around
- 1995 and decrease slightly in the following .five y.ears. Only a 19% increase
in total usage is projected between 1980 and the year 2000. This corresponds
to a 38% increase in particulate emissions between 1980 and 2000 based on a
projected 89% increase in stove wood usage and a 9% decrease in fireplace
£
wood usage. These projections are for residential wood combustion (RWC)
f *
| within Seattle limits only.
fc.-
For residences within the Spokane city limits, total wood fuel use is
L projected to increase by 4% between 1980 and 1985. A 7% increase in total
r wood usage is projected between 1980 and the year 2000. RWC particulate
Li
r
-------
emissions are projected to increase by 26% between 1980 and the year 2000
based on a 93% increase in stove wood usage and a 19% decrease in fireplace
wood usage.
These long-term wood fuel use projections represent most probable
(base case) input data. No alternative input data or assumptions have been
tried other than model calibration runs and limited sensitivity testing.
Additional model runs could test (1) alternative input data, e.g., revised
projections of electricity costs under the Northwest Electric Power Planning
and Conversation Act of 1980 (PL 96-501)-; or (2) effects of certain RWC
control strategies, e.g., weatherization to lower average home heating
requirements or tax credits for installing cleaner wood stoves, etc.
The future prices of both wood and the conventional fuels have major
impacts on the projected levels of residential woodburning. For conventional
fuels, projected prices were furnished by the Bonneville Power Administration
and the Oregon Department of Energy. These cost projections show oil, gas,
and electricity approximately doubling in real cost between 1980 and 2000,
compared to about a 50% increase in real cost for wood during the same period.
Changes in these projected costs, as well as the possible substantial increase
in residential coal combustion, could alter the wood use projections sig-
nificantly. As an example, changing the wood price increase from +2Vyear
to +5%/year results in a net decrease in residential wood combustion. These
projections should be used with the understanding that major deviations in
fuel costs will have a substantial impact on the levels of residential wood
combustion.
xii
-------
I. INTRODUCTION
A. PURPOSE
The primary purpose of Task 3 is to project trends in residential wood
fuel use through the year 2000 for three cities in the Pacific Northwest:
Portland, Oregon, and Seattle and Spokane in Washington. Projections for
the years 1985, 1990, and 1995 will be estimated in the process of formulating
a best estimate for the year 2000. Shorter term trends, based on data
describing recent wood fuel use and othgr trend factors, will also be examined
where sufficient data is available.
The projected trends in residential wood fuel use from Task 3 will be
used in Task 2A to project future ambient air quality impacts from residential
wood burning. The projections should assist air quality agencies in estimating
the severity of future impacts from wood burning. The modeling approach used
to project long-term trends will also indicate how important factors can inter-
act in determining future wood burning impacts, such as the estimated price
of wood versus conventional fuels.
B. SUMMARY OF TECHNICAL APPROACH
A thorough literature review was conducted to locate information on
previous projections of residential wood fuel use. Extensive contact was
made with knowledgeable individuals in public agencies, private institutions
and businesses, universities, and associations. Most projections were of
national wood fuel use. Transportation costs are a major factor in residential
wood combustion, which tends to make national trend data of minimal use in
predicting trends for specific localities. For example, two cities only 200
1
-------
miles apart can have very different residential wood burning patterns because
of the differences in wood availability and cost.
Only two projections were for Pacific Northwest areas - one by the Bonne-
ville Power Administration (BPA) and the other by the Oregon Department of
Environmental Quality (ODEQ). The BPA study covered only electrically heated
homes in the Pacific Northwest. The ODEQ study used survey results from the
1978-79 heating season in the Portland-Vancouver area, and projected residential
wood combustion through 1987 using expected population growth and trend data
for wood cutting permits.
A particularly relevant study of residential wood fuel use in New England
was recently completed at Dartmouth College. It provided the best critique
of previous projections of wood fuel use in the U.S. The Dartmouth study also
provided a methodology (model), which was adapted and used in this study to
project wood fuel use trends for the Portland, Seattle, and Spokane areas.
The model's input data and assumptions take into account those estimated
costs of wood and alternative fuels, population and household growth projec-
tions, household heating requirements, area mix of home heating fuels, and
other factors influencing residential wood fuel use.
Pacific Northwest input data for these and other model variables
included the following:
historical and projected costs for conventional fuels such as
oil, gas and electricity
historical and projected fractions of households that rely on
oil vs. gas or electricity as primary heating fuels (used to
derive average conventional fuel heating costs)
historical and projected household heating requirements
past and projected wood costs (used to assess savings potential
from heating with wood compared with conventional fuels)
-------
assumed wood heating installation costs
assumed inconvenience costs of wood heating
assumed rates of consumer responses to install wood heating
systems based on the expected length of the investment pay-
back period
The model's estimates for the years 2000, 1995, 1990 and 1985 are
presented as this study's "best estimates" of long-term wood fuel use
trends. A tentative estimate of the uncertainty in the model projections
is made based upon a limited sensitivity analysis of major variables. Model
projections are compared (in Section III) with the national wood use trends,
the two previous wood use projections in the Pacific Northwest by BPA and
DEQ, and regional population and household growth projections.
Short-term trends for the three Pacific Northwest cities were estimated
using several different types of information, including:
1. Direct measures of firewood use, including volumes of firewood
removed from public and private lands under firewood cutting
permits and household surveys of firewood use.
2. Direct measures of RWC effectsspecific ambient air quality
impacts of RWC represented by nephelometer data.
3. Other data sources considered but not used in trend calculations
included data on the sales of wood burning appliances and other
census and air quality data. Air quality data used was normalized
for heating season severity (based on heating degree days) and
meteorology (based on surface wind speed). Trend data were plotted.
Regression lines of best fit for these plots were used to describe
annual rate of change in trend parameters from 1982-85 for short-term
trend estimates.
Previous chemical mass balance and carbon-14 analyses of particulate
monitoring samples, which are considered reliably specific indicators of the
ambient air quality impacts from residential wood combustion were examined.
There were not enough of these to do trend analysis. However, their magnitudes
and location are discussed.
3
-------
The short-term trend estimates confirm what was already known -- that
residential wood combustion has increased substantially in the last few years
However, the long-term trends projected for wood fuel use are likely to be
more reliable bases for projecting air quality impacts from residential wood
burning for several reasons:
* Penetration into the marketplace of a new technology, almost
always follows an S - shaped curve with increasing penetration
rates at first. Depending on the stage of penetration, short-
term trends are generally not good long-term trend factors.
Short-term trends can be erratic, reflecting differences in
consumer behavior, dips and swings in the economy, meteorological
di fferences.
Air quality 'strategies normally need to look at a long-term
perspective to attain and maintain air quality standards.
The long-term projection methodology utilized in this report
projects how key factors influencing wood use are likely to
change and incorporates those changes in the projections.
Such key factors include projected growth in households,
projected fuel prices, projected level of reliance on oil vs.
gas vs. electricity for household heating, and others.
C. REPORT OVERVIEW
The remainder of this report is organized as follows: (a) Section II
describes this study's short-term projections of wood fuel use for the three
cities: Portland, Seattle, and Spokane; (b) Section III contains long-term
trend projections through the year 2000 based on adaption of the Dartmouth
model to the three Pacific Northwest cities; and (c) Section IV lists the
references and personal contacts that provided the information utilized in
this study.
Appendix A contains detailed documentation of the values which were
input into the long-term trend model (e.g., household heating requirements,
-------
historical and projected fuel prices), and the model changes that were made
to apply the projection model to the Pacific Northwest. Appendix B contains
a discussion of some of the factors not included in the long-term trend model
-------
II. ANALYSIS OF SHORT-TERM WOOD FUEL USE TRENDS
The overall objective of this analysis of short-term wood fuel use trends
is to quantify recent trends in wood fuel use for the three cities and to
project how this may change over the next few years. For purposes of this
analysis, "short-term" will refer to trends extended through 1985. Section
III estimates long-term trends through the year 2000.
To estimate wood fuel use trends, it would be ideal to have a multi-year
data base that quantifies the area-wide-total amount of firewood used in homes.
This is not available. What is available is a number of direct and indirect
indicators, which describe an unknown fraction of the total firewood usage or
some indirect measurement of firewood use rates (e.g., sales of wood stoves),
or direct measures of the effects of RWC (e.g., air quality impacts from RWC).
Some of the parameters do not "cleanly" measure RWC because of interfering
influences. All have inherent uncertainties that are difficult to quantify.
Since no single available indicator adequately estimates the magnitude
and trends in home wood fuel use, a number of different indicators are r.eviewed
here. Their combined evidence is used to derive a "best estimate" of short-
term trends (through 1985). The short-term trend parameters considered are
described and evaluated in Section A. Section B describes the findings and
conclusions of the short-term trend analyses using the selected indicators
with documentary details in Appendix A. Section C summarizes the best quantita-
tive estimate of short-term RWC trends for the three cities: Portland,
Seattle, and Spokane.
-------
A. SHORT-TERM TREND PARAMETERS CONSIDERED
In analyzing wood fuel use trends, more weight should be given to para-
meters that directly measure home firewood use than to parameters that
indirectly measure wood use or directly measure its effects, e.g., its air
quality impacts. The short-term trend parameters considered in this study
are listed below in these three categories:
1. Direct measures of firewood use
a. Volume of firewood removed from public and private lands
under firewood cutting permits
b. Household surveys of firewood use
2. Indirect measures of RWC
a. Sales of wood burning appliances
3. Direct measures of RWC effects
a. Ambient air quality impacts related to RWC
These parameters are described and evaluated in the following subsections.
Several criteria were established in advance for the types and amount of data
that should be available. First, at least three consecutive years of recent
data should be available -- including the three most recent heating seasons.
(The six months from October through March were considered a heating season.)
Second, no major data gaps should be present. For example, gaps in hourly
air quality data might not be significant because the data would eventually be
averaged over a heating season for trend analysis, but several missing weeks
of data would be significant. Third, the data must be judged reliable.
Quantifying Firewood Use
One of the two most direct measures of wood fuel use trends would be to
-------
quantify, over a series of years, the total amount of firewood used in a given
metropolitan area. Two ways to do this are: (1) to monitor all of the sig-
nificant sources of firewood supply in an area e.g., firewood removed from
public or private lands under wood cutting permits; or, (2) to determine area-
wide household firewood use, using survey techniques.
The major sources of firewood for RWC are state and federal forests,
privately owned forests which allow firewood removal, and wood processing
facility wood waste. Firewood supplies for different metropolitan areas are
made up of various mixes of these major_supply sources, depending primarily
upon land ownership patterns in the region, timber management practices, and
market conditions.
Tracking RWC trends by firewood use requires data over a series of
years or heating seasons based upon the total amount of firewood used in a
given region. This type of information could come from quantitatively
monitoring all major sources of firewood supply or from statistical areawide
surveys of household firewood consumption. Available information of both
types is discussed below.
Information on Volumes of Firewood Cut for Home Use
People either cut their own wood or buy it from a commercial wood
cutter or retailer. Fifty to seventy percent of the respondents to the Task
2B surveys cut their own. The increasing need for permits by individuals and
commercial cutters obtaining wood from public or private land provides one of
the only data bases on the volumes of firewood used in an area.
Wood cutting permit data typically afford only a rough approximation of
the total areawide firewood use. While permits issued to individuals by
-------
public and private land managers often specify a limit on the amount of wood
that can be taken (typically 5-10 cords/permit), generally no one checks to
see how much is actually taken.
Commercial wood cutters usually operate under a stricter permit system
or timber sale contract, which does account for the volume of wood taken for
purposes of payment. Commercial wood cutters may also subcontract with
logging contractors to remove some of their logging residues. Public timber
sale contracts give logging contractors first rights to all logging residues
or slash. The logging contractor must .agree on a plan of disposing of the
slash, either by approved burning or other removal, to facilitate reforestation
of the site. Contractors are often required to yard unused materials (YUM)
to piles near roadways. Such YUM piles become a prime source of firewood and
are offered for public consumption under wood cutting permits if the con-
tractor does not otherwise dispose of it.
A growing practice mentioned by a number of forest staff involves truck
loads of large logs delivered to groups of homeowners, who may pay over $500
per truck load. This relieves homeowners of the transportation costs and some
of the inconvenience of obtaining firewood. Typically only a nominal fee
($.50/cord) for such removal of cull logs or logging residues is charged,
but forest staff believe many of these truckloads are not reported in any
permit system.
More private landowners, such as timber and pulp/paper companies, are
also establishing permit systems to allow the public or commercial cutters to
remove logging residues in their forests or unwanted wood from "sort yards"
at mill facilities. The permit system also serves to control access to private
wood supplies.
9
-------
The following subsections summarize information about firewood volumes
available from woodcutting permits issued for public and/or private lands
in the vicinity of Portland, Seattle, and Spokane.
Portland. Oregon
The Mount Hood National Forest is by far the major supplier of firewood
to the Portland area. Few state or private forest lands exist near enough to
supply Portland with much wood. Much of this land is in second growth timber
with little active logging.
Table 2 summarizes the estimated ruimber of wood cutting permits issued to
2
the public in recent years from the Mt. Hood National Forest . Permit totals
were converted to firewood volumes using an estimate by Mt. Hood National
Forest staff that an average of 3.5 cords were removed per permit. The
staff did not perceive any slackening in the escalation of firewood demand
that has been going on for several years. However, they did foresee stronger
competition for available logging and wood processing residues from several
new private enterprises planned or under construction, including two wood
densification plans and several wood-fired electric power producing operations
(cogeneration). The demands for raw wood from these facilities could strongly
affect the availability, accessibility, and cost of firewood in the Portland area.
Seattle, Washington
Seattle area residents obtain significant amounts of firewood from federal,
state and private sources. Table 2 summarizes information on wood cutting
permits issued for the Mt. Baker/Snoqualmie National Forest, extensive state
forest lands operated by the Dept. of Natural Resources (DNR), and several
sort lots and forest lands near Seattle operated by Weyerhaeuser, Inc.
10
-------
TABLE 2
Estimated Wood Cutting Permits and Firewood Volumes
Removed from Selected Federal, State and Private Sources
near Portland, Seattle and Spokane
City
Permit Source Number of
Year Permits Issued
A. Portland, Oregon
1. Mount Hood National Forest2a
1976 12,000
1977 16,000
1978 24,000
1979 30,000
1980. 39,000
1981 48,000
B. Seattle, Washington
l.a. Mount Baker/Snoqualmie National
7/75-6/75
9/76-S/77
9/77-8/78
9/78-8/79
9/79-8/80
9/80-8/81
l.h. Mount Baker/Snoqualmi e National
cutting)
1979
1980
19S1
Volume of
Board Feet
(xlOS)
Forest 2b
3.629
4.769
5.518
12.265
16.081
16.157
Forest2b (
.255
.700
.710
Firewood Removed*
Cords/Year
(xlQ3)
42
56
84
105
137
168
(general public)
7.5
9.5
11.0
24.5
32.2
32.3
'4 Districts/commercial
0.51
1.40
1.42
Volumes provided in board feet were estimates by forest officials.
Volumes in cords were calculated using (1) number of permits x average
cords taken/permit as estimated by forest staff; or (b) from board
feet estimates, using 1 cord = 500 board feet.
11
-------
TA3LE 2 (continued!
Citv
Permit Source Number of
Year
2. ',,'ashi
1978
1979
1930
1931
Permits Issued
nqton Department of Natural
(3900)
(3000)
2855
1852
Volume of Firewood
Board Feet
(x!06)
Resources^
Removed*
Cords/Year
(xlQ3)
509 (98. 23)**
2,170 (99.6%)
835 (99%)
110 (95%)
3. Weyerhaeuser, Inc.^
1975
1975
1977
1973
1979
1930
1981
C. Soo'one,
1. Colvi
1977
'1978
1979
1980
1981
2. Idaho
1977
1978
1979
1930
1981
760
920
1,500
3,534
7,350
10,500
19,500
Washinqton
lie National Forest2e, 2f
3,128
'4,071
4,845
9,810
9,818
Panhandle National Forest*^
1,140
1,250
2,950
' 3,254
3,535
(?)
15.7
24.7
25.7
41.5
32.1
3.08
3.40
7.95
8.78
9.81
(?)
31,400
49,400
53,800
83,200
64,200
6,156
6,804
15,930
17,571
19,629
* Volumes provided in board feet were estimates by forest officials.
Volumes in cords were calculated using (a) number of permits x average
cords taken/permit as estimated by forest staff; or (b) from board
feet estimates, using 1 cord = 500 board feet.
**Values in parentheses indicate percentage of total which was taken
by commercial cutters.
12
-------
O I
Mt. Baker/Snoqualmie staff provided estimates of the firewood removed
(million board feet) under wood cutting permits by the general public
(estimated at about 3.25 cords/permit in 1980-81) and for commercial wood
cutters in four forest districts. Both estimates show a marked leveling in
1980-81 compared to the large increases in recent years. The reason for this
apparent leveling of demand was not known. Reduced logging activity due to
the depressed homebuilding market reduced normally available volumes of logging
residues by over 50%.
Extensive state (DNR) forest lands_are near Seattle . Many are closer
than national forest lands and provide easier access (less rugged). DNR's
toll free phone receives many request (-25) each day for firewood information.
In response to continuing strong public demand, the State Forester has requested
staff suggestions on making more firewood available to the public.
The DNR wood cutting permits and firewood volumes removed (Table 2,
part A2) show an unusual pattern of sharp decline in the last few years. A
combination of heavy commercial cutting and the present decline in logging
activity may have temporarily depleted wood residues available for firewood.
Commercial cutting under DNR timber sale contracts accounted for 95-99+% of
the firewood volumes removed from these DNR lands, which supply far more fire-
wood to the Seattle area than the Mt. Baker/Snoqualmie National Forest. How-
ever, careful permit records have only been kept for the last two years.
Weyerhaeuser permit totals refer to both cutting on Weyerhaeuser forest
p j
lands and in sort yards at Weyerhaeuser wood processing facilities . A
change from 2-4 week permits to daily permits in 1979-80 probably accounts
for a significant portion of the increase in the total number of permits
issued in recent years and makes this data questionable for trend purposes.
13
-------
Spokane, Washington
Both the Colville and Idaho Panhandle National Forests are near enough
for firewood cutting by Spokane area residents. Table 2 includes wood cutting
permit totals and/or forest staff estimates of firewood volume removed from
2e
both forests . The 1981 data from Colville National Forest show a leveling
2f
of wood cutting permits and a substantial decrease in firewood volume removed
However, this may be due to a change in permit procedures, from personally
issued by Forest Service staff to self-issuing permits. The change may also
have influenced staff estimates of the average number of cords/permit removed.
A written survey of 1981 permit holders was the basis for a staff estimate
that an average of 5 cords of firewood were taken under self-issued permits.
State forest lands provided Spokane residents with an estimated 4000-5000
O L
cords per year in 1981, according to DNR staff , who said total firewood
volumes removed continued to increase approximately 15% per year. Reliable
information is hard to obtain because a public wood cutting permit system was
not implemented until 1980.
Overall, wood cutting permit information has serious limitations as a data
base for estimating wood fuel use trends. It is not complete or reliable enough
to represent the actual total firewood use in an area. Much firewood can be
obtained outside the permit system, and the permit systems themselves do not
monitor volumes of firewood actually taken. Trying to inventory the total
amount of firewood used in a metropolitan area over successive years to
establish a trend would require relatively complete accounting for firewood
volumes obtained from all major public and private landholdings, plus an
estimate of wood obtained elsewhere -- a formidable task indeed.
14
-------
Individual permit programs can provide a glimpse of what is happening in
a sector of the firewood supply system - especially if reasonably accurate
information is available. Analysis of a number of these permit systems in
different locations in a region show a pattern of wood fuel use. The data in
Table 2 show different patterns in Portland (continuing increase in permits)
versus Seattle and Spokane (leveling of permits).
Wood Cutting Permit Trends
Figure 1 shows plots of selected data intended to represent trends in
major portions of the total amount of firewood supplied to households in the
three metropolitan areas. Estimated firewood volumes removed under cutting
permits (cords/year) during the last 4-6 years are plotted. Dashed trend lines
are regression lines of best fit.
All four trend lines show a notable similarity, exhibiting a 5.7-7.9%
average annual increase in firewood volume over the last 4-6 years. Trend
lines, which are less steeply sloped, begin with smaller initial firewood
volumes. Incidentally, this approximate 6-8% annual rate of increase agrees
will the model simulation of RWC activity during this same time period
(Section III).
For two areas (Seattle, Spokane/Colville National Forest), the most recent
(1981-82) firewood volumes are plotted, but are not included in the regression
line calculation because they deviated substantially from the trend of the
previous few years. In these cases, for the most recent heating season, total
wood cutting permits and/or firewood volumes remained about the same as the
previous (1980-81) season. In one case (Colville National Forest), estimated
firewood volumes removed were down 25% from the previous season. This may have
15
-------
150,000 ..
100,000 ..
50,000
1975
1980
1985
<£
Figure 1. Recent Trends in Firewood Volumes Removed Under Wood Cutting Permits
from National Forests Near Portland, Seattle and Spokane.
16
-------
been due to a change in the methods of estimation used by forest officials.
In any event, the results of these trend analyses indicate that firewood
supplied by federal forests near the three cities have been increasing over
the last 4-6 years at an annual average rate of about 6-8%, but there are now
preliminary indications of a leveling of this trend in at least several of
these areas. Aside from the homebuilding/logging slump, which has decreased
the amount of logging residues available in most forests, forest officials
were not able to identify any decrease in demand for firewood for RWC. Most
officials felt demand was strong and wow-Id continue to be so.
Statistical Surveys of Wood Fuel Use by Residence
Survey techniques would appear to be a simpler and more accurate way to
track total firewood consumption in a metropolitan area than attempts to
inventory firewood cutting permit information, etc., provided such surveys are
designed to represent the entire area. Survey data over successive heating
seasons would be necessary for trend analysis of areawide wood fuel use.
The Oregon Department of Environmental Quality (DEQ) conducted a telephone
survey in. 1979 to determine the amount of RWC in three metropolitan areas
(Portland-Vancouver, Eugene-Springfield, and Medford-Ashland) during the 1978-
79 heating season . This information was used to characterize residential wood
space heating as a source of particulate emissions for air quality control
planning purposes. Base year (1978) RWC emissions were projected to grow based
on either population/household projected growth rates or a more rapid increase
using a trend factor based on recent increases in timber harvest volumes. The
survey determined the number of cords of firewood used by an "average household".
Based on the total number of households in each area, the area total emissions
were calculated.
17
-------
Area total firewood use was not calculated, but is derived from the data
collected. For Portland-Vancouver, the average household was estimated to
have used 0.89 cords in 1978; 1.73 cords per average wood burning household
No wood was burned in 45.5% of Portland households . Thus, Portland-Vancouver's
estimated (1978) total of 428,600 households a burned approximately 360,000
cords of firewood in the 1978-79 heating season. This is considerably more
than the firewood estimated to have been supplied from the Mt. Hood National
Forest (Table 2) in 1978 or 1979.
DEQ surveyed random samples of 400-households in each metropolitan area.
The survey was designed to be statistically reliable within ±5% and representa-
tive of the entire Portland(OR)/Vancouver(WA) metropolitan area. No other
comparable surveys have been done for Portland. Similar data which might
support trend analysis of metropolitan firewood consumption is available for
other years. Based on this survey, using the two growth factors mentioned
above (population/household growth and growth in timber harvest volume wood
cutting permits), DEQ projected a growth in total metropolitan firewood
consumption of 139% between 1977 and 1987. This implies total metropolitan
firewood use of 740,000 cords/year by 1987.
The Task 2B survey of a Portland area residential neighborhood found that
14.2% of its 1981 households used wood as the primary heat source (28% of all
households had wood stoves), and 54% of all households burned some wood. The
average number of cords burned during the 1981-82 heating season was 2.0.
However, these figures are probably not representative of the Portland metro-
politan area as a whole because this survey forcused on a single residential
neighborhood chosen for apparent high levels of residential wood burning.
18
-------
Neither the Seattle nor Spokane areas had random surveys of areawide
wood use. Existing survey data for these two cities ' ' including the
Task 2B neighborhood surveys, are discussed in the Appendix A (Section&2D and
3A(1), respectively). The assumptions made in using this survey data to
estimate wood fuel use by residences are explained in these Appendix sections:
Until some comprehensive system of tracking residential wood combustion
over a series of years is implemented, trend analysis of wood fuel use will be
limited by data series which represent an unknown fraction of total firewood
usage. The most practical methods of quantifying areawide and household wood
use involves statistically sound random surveys.
Indirect Indicators of Wood Fuel Use
Sales of residential wood burning appliances and census data were examined
for possible use in estimating wood fuel use trends. Neither trade associations
nor manufacturers contacted could or would release actual sales information of
use in estimating RWC trends.
Using Bureau of Census data, Cooper plotted the national total number of
23
.wood burning stove type devices shipped in this country since 1960 . This
showed a sharp increase in stoves shipped beginning in 1973 (Figure 2). New
23
data from the U.S. Dept. of Commerce will soon be available to extend this
plot beyond 1979. It is possible to obtain only the data for the Pacific
Northwest. However, this data was not available within the time and resources
of this study. The Oregon and Washington stove sales data represent units sold
from those states, which can include a significant amount of out-of-state
deliveries. Thus, this data does not represent wood fuel use in the Pacific
Northwest so much as sales activity.
19
-------
Figure 2. Twenty Year Trend in the Number of Wood-burning Stove Type
Devices Shipped in the USA.
0 o
>- 10
O
-------
Moreover, while such data on stove shipments contribute qualitatively to
the overall picture of recent sharp increases in RWC activity, they do not
provide direct quantitative estimates of wood fuel use, since the amount of
wood used per stove is unknown. Accordingly, no analysis of sales data was
done to quantitatively estimate short-term RWC trends.
Comparison of 1970 and 1980 census data related to RWC could theorectically
shed some light on RWC trends during that ten year period. However, census
data provides very little information to estimate wood fuel use or related
trends.
The type of fuel used to heat residences is obtained from a census
questionnaire, but this information for 1980 will not be published until late
1982 in "Detailed Housing Characteristics". No census information is obtained
on secondary (supplmentary) heating equipment of fuel. The Bureau of Census
data on the number of wood burning stove type devices shipped in this country
was discussed in the previous section. Accordingly, no trend analysis of
census data was pursued.
Direct Indicators of Effect of Wood Fuel Use
Residential wood fuel combustion contributes substantially to ambient
24 25
air quality problems ' . Of primary concern are RWC contributions of
respirable particulate and carbon monoxide (CO) levels. Thus, the following
types of air quality monitoring data related to respirable particulates or CO
were considered to determine if they could provide a continuous series of data
whose trends might be associated with residential wood fuel use trends:
1. Carbon monoxide monitoring data
2. Certain regularly collected air quality monitoring data
related to respirable particulate levels, including:
21
-------
a. Light scattering Coefficient (Bscat) -- The integrating
nephelometer, measures the degree to which small particles
[, in the air scatter light, indicated by an hourly average
scattering coefficient (Bscat). RWC emissions are largely
(90%) small particles.
b. Coefficient of Haze (COH), or Soiling Index -- Air filtered
through a tape produces a spot whose darkness is measured
I" by light transmission, indicated by an attenuation co-
fi efficient measured every two hours.
c. Benzene soluble particulate extractions -- Benzene extract-
able non-polar organic materials are associated mainly
with combustion process (like RWC), auto exhaust, industrial
emissions, and secondary organic aerosols. Gravimetric
analysis of extractions gives total "organic" carbon in
particulate samples. _
d. Total carbon levels and organic carbon to total carbon
ratios in particulate (Hi-Vol) samples -- Both these
parameters have been shown to rise substantially during
winter months, presumably due to combustion of carbonaceous
fuels.
Other Air Quality Data
Some air quality sample analysis techniques, for example chemical mass
j balance (CMB) and Carbon-14 analyses of fine particulate samples, directly
t
estimate RWC impacts and non-fossil carbon impacts, respectively. The
L "modern" (non-fossil) carbon impacts can occur from sources besides RWC,
H
I
such as slash burning, veneer dryer emissions, sanderdust and plant materials
(pollen, etc.).
Only the light scattering (Bscat) data was eventually analyzed for trends
in this study. The other types of data were considered to be either too
subject to interfering sources or had problems with collection efficiency and
general accuracy. In some cases, there was simply not enough continuous data
to support trend analysis.
f 22
[
[
-------
While evening peaks in ambient CO levels may correlate well with RWC
activity, separation of vehicle contributions to CO levels from RWC con-
tributions makes it impossible to use CO data for trend analysis.
Coefficient of Haze (COH) data suffers from poor collection efficiency
of paper tape samplers and general inaccuracy of measurement method. It also
reflects impacts from sources other than RWC, such as diesel exhaust, residual
oil burning, and fine particulate from industrial emissions.
Benzene soluble extractions of total particulate samples were only
available for Portland. DEQ discontinued benzene extractions several years
ago. No data is available for the last several heating seasons when the
sharpest increase in RWC occurred.
Trend analysis was not performed on total carbon analyses of particulate
samples or organic/total carbon ratios for several reasons. Total carbon
levels can be influenced by many sources, including auto and truck exhaust,
RWC, plant tissues, sawdust, spores, pollen, etc. Even winter carbon levels,
typically elevated presumably by combustion heating sources, would at best
represent an upper bound for RWC levels. Discontinuance of benze%ne extraction
several years ago in Oregon meant the only data for organic/total carbon ratios
available did not cover the most recent heating seasons.
Chemical Mass Balance(CMB) analysis has shown that RWC can be one, if not
the major contributor to respirable (<2 microns) and inhalable (<10 microns)
particulate levels in Portland and Medford, Oregon. An increasing number of
fine particulate monitoring sites are being operated on a regular basis in the
Pacific Northwest, but none has accumulated enough continuous data for trend
analysis. In the future, periodic measurement of respirable or inhalable
23
-------
particulate mass and CMB analysis of RWC contributions to it, is likely to be
the best indicator of RWC trends. However, trend analysis using CMB data
would require a large number of periodic samples judged to be representative
of winter season air quality.
Task 2A reviews past maximum fine particulate levels observed in Portland,
Seattle, and Spokane and several other cities in the Pacific Northwest, including
available CMB analyses of RWC levels. These historical samples and CMB analyses
provide valuable information on RWC levels at various times and places, but this
information is not plentiful or continuous enough for trend analysis.
Carbon-14 analyses of fine particulate samples collected during heating
26 27
seasons can be one of the most specific analyses of RWC ' ' RWC emissions
contain entirely "modern" carbon, in contrast to fossil fuels whose greater age
1 A 1O
has resulted in more C1 decay to the isotope C . Thus, the fraction of a
particulate sample contributed by "modern" carbon sources is determinable
through measurement of 12C/14C ratios, which can be sensitively and accurately
measured. Provided a suitable fine particulate sample is collected, which
excludes larger particle size sources such as raw wood fiber or other plant
materials, the "modern" carbon component can be attributed to a very limited
number of sources, such as RWC or slash burning. Then for major urban areas,
other measurements like background site particulate levels and nephelometer
data can be used to ascertain whether slash burning smoke intrusions were
occurring. On sample days when other vegetative burning influences, like slash
or open burning, are known not to have occurred, the vegetative burning impact
can be attributed to RWC. However, relatively few carbon-14 analyses have been
done to date due to the expense (approximately $1000/sample), and a determination
of trends is not possible with the limited available data.
24
-------
Trends in Light Scattering (Bscat)
Light scattering (Bscat) measurements can also be influenced by
sources of very fine particulate other than RWC - e.g., forest fires, grass
field burning, auto traffic, industrial upsets, and secondary aerosol forma-
tion. Meteorology also directly influences measured Bscat levels. Several
precautions were taken to screen out or eliminate such interferences prior
to trend analysis of Bscat data. ,N evetheless, the results are presented
with the strong caution that they represent trends in ambient air quality
impacts which may be associated with RWC.
Only Bscat data collected during a "heating season" (here defined as
October - March; 6 months) were analyzed to maximize the influence of RWC and
eliminate the seasonal effects of such activities as forest fires and grass
field burning. Bscat data were normalized for temperature (using heating
degree days) and wind speed, to minimize variations due to meteorological
factors, as follows:
Measured Monthly Ave. Wind Speed Monthly Ave. Bscat
Monthly Average X "Normal" Monthly Ave. Wind Speed = Normalized for
Bscat Monthly Total Heating Degree Days Temperature and
- "Normal" Monthly Total Degree Days. Wind Speed
Normalization factors were developed using wind speeds and heating degree
days measured at the nearest National Weather Service site. "Normal" values
typically representing average values over a period of 30 years or more, were
also obtained from National Weather Service records.
Monthly average normalized Bscat data were further averaged to represent
a heating season (October - March) and plotted versus time for trend analysis.
Visual inspection and regression lines of best fit were used to assess trends
25
-------
Table 3 summarizes the Bscat data analyzed. Figure 3 shows the resulting
plots and (dashed) regression trend lines of best fit.
The sites from which data were selected for analysis were chosen primarily
because they provide the longest series of continuous data including the three
most recent heating seasons. Site selection was limited. Most sites were
surrounded by more commercial or industrial rather than residential land uses.
No site in Spokane had the requisite minimum data requirement - i.e., data for
the three most recent heating seasons. Portland had sufficient Bscat data for
both a downtown site and a rural (background) site.
Figure 3 shows an increase in "heating-season-average" Bscat values at
the Seattle and Portland sites of 6-8% per year over the three most recent
heating seasons. The average annual increase is almost 8% for the Seattle (Kent)
site, and almost 6% at the two Portland area sites. The downtown Portland sites
exhibit this rate of increase in normalized Bscat data for the last eight years.
The background site values closely track the downtown site values for years
where data was available at both.
B. FINDINGS AND CONCLUSIONS
The results suggest that respirable particulate air quality during the
heating season has been deteriorating in Portland and Seattle at a very
noteworthy rate in recent years. Emission inventory projections and computer
modeling results for 1977 vs. 1987 in the Portland Particulate SIP indicate
that fine particulate impacts from other major sources are likely to decrease
or only increase slightly. Therefore, the increases in Bscat levels are
believed attributable to increases in RWC.
26
-------
TABLE 3
Summary of Short Term Trend Parameters Quantified
Short Term
Trend Heating Seasons
Parameter . Trended Estimated Annual Ave.
City (No. of Seasons) Rate of Increase (%)
A. Firewood Volume Removed from
Nearest National Forest(s)
1. Seattle, Washington
Mt. Baker/Snoqualmie N.F. 1975_-80 (5) 7.8-7.9
2. Portland, Oregon
Mt. Hood N.F. 1976-81 (6) 6.5
3. Spokane, Washington
Colville N.F. 1977-80 (4) 5.7
4. Spokane, Washington
Idaho Panhandle N.F. 1977-81 (5) 7.2-7.3
B. Heating Season - Average Light
Scattering Coefficient (Bscat)
1. Seattle, Washington'7
Kent Site 1978-81 (3) 7.9
2. Portland, Oregon8
a. CAMS Site 1978-81 (3) 57
b. Carus Site 1978-81 (3) 5^5
27
-------
c
-s
ro
Nephelometer Light Scattering Coefficient (Bscat
m
-1
ro
oo
~O 73
O fD
T O
r+ fD
OJ c-t-
3
0. -i
~5
OJ ro
CX CX
C/1
ro -"
OJ 3
o
ro ~i
r+
ro
ro
ex
2
n>
X3
3-
ro
o
ro
r*
ro
T
O
OJ
o
OJ
CU
ro
ro
-j
OJ
rv
O
^
-------
At a 6-8% annual rate of change, the average pollution level would double
within 9 to 12 years. The recent rate of increase in Bscat is the same as that
estimated above for firewood supplied by federal forests. This agreement lends
weight to the obvious association between this period of deterioration in
respirable particulate (Bscat) air quality and an increase in RWC activity.
Further support for this association comes from chemical analysis of fine
particulate samples, which shows that RWC is a major contributor to respirable
particulate 1 eve!s.
C. BEST ESTIMATE OF SHORT-TERM WOOD FU?L USE TRENDS
Table 3 indicates that for the two short-term trend parameters for which
calculations were made - i.e., firewood removed from nearby federal forests,
and heating-season-average Bscat data - the recent annual average rate of
increase was 6-8%. Assuming RWC activity has increased in proportion to these
related trend factors, our best estimate of the short-term trend in residential
wood fuel use is a 6-8% maximum annual average increase. At this annual rate
of increase, wood fuel use would increase 34-47% between 1980 and 1985.
Recent indications of a leveling of firewood supplied by several federal
forests may signal that the recent trend has begun a significant slowdown.
The long-term trend projections in Section III of this report indicate that
wood fuel use will grow at different rates in the three cities between 1980
and 2000. Between 1980 and 2000, Portland and Seattle are projected to have
annual average growth rates in total wood combustion of 2.0% and 1.0% per
year, respectively. Spokane is predicted to have a net increase averaging
0.3% over this twenty year period. However, as discussed in the Findings and
29
-------
Conclusion of the Long-Term Trends Section, emissions are predicted to increase
proportionally more in all areas due to the switch from burning in fireplaces
to burning in stoves which have a greater emission rate.
30
-------
III. ANALYSIS OF LONG-TERM WOOD FUEL USE TRENDS
INTRODUCTION
Long-term projections for use of a resource are generally difficult
because factors unforeseen at the time of projection often become important
influences later and can radically alter the expected behavior. If, for
example, the task of projecting residential wood usage had been undertaken
in 1970, it is doubtful that many projections would have foreseen the influence
rising prices for oil, natural gas, and electricity have had in motivating
households to convert to heating their dwellings with wood. Nevertheless, long-
term projections of residential wood usage in the Pacific Northwest are of
interest in estimating its severity on future air pollution problems. The
analysis described in this section projects residential wood usage through the
year 2000 (at 5 year intervals beginning with 1985) for the cities of Portland,
Seattle, and Spokane. Portland projections were made on an AQMA basis (roughly
equivalent to the Standard Metropolitan Statistical Area), rather than on the
city itself, since the baseline information on wood fuel use in 1978 was
available for the larger geographical area.
The remainder of this section is organized as follows:
Methods for Long-Term Wood Fuel Use Projections (Section A)
Description of Chosen Analytical Approach (Section B)
Marshall's economic model
Addition of a fireplace wood usage sector to the model
Model Calibration
Model Sensitivity Analysis
Findings and Conclusions (Section C)
Best Estimate of Long-Term Wood Fuel Use Projections (Section D)
31
-------
Specific details on factors input into the projection model are presented
in Appendix A.
A. METHODS FOR LONG-TERM WOOD FUEL USE PROJECTIONS
A wide variety of information sources were examined in order to seek
projections within the literature for residential wood usage:
Library systems at the Bonneville Power Administration and the
Western Solar Utilization Network (SUN) were completely reviewed
for specific projections of wood usage and general methodologies
for projecting energy usage.
Federal, state and local agencies in Oregon and Washington were
contacted such as the Oregon Department of Energy (ODOE), Washing-
ton Department of Natural Resources (WDNR), Washington Department
of Ecology (WDOE), Oregon Department of Environmental Quality (ODEQ),
U.S. Forest Service (USFS), and others. No year 2000 projections
were available. Only the ODEQ has short-term wood usage projections
which are described in Section II A.
A search of EPA computerized bibliographies was conducted for
subject matter relating to residential wood burning.
Air Pollution abstracts were searched for wood related articles
for the last three years as was the magazine Forest Industries.
Wood stove and fireplace related trade associations such as the
Wood Energy Institute in Portland and the Wood Heating Alliance
in Washington, D.C. were contacted.
Information from such sources is referenced in this text where appropriate,
and a comprehensive list of references is presented in Section IV.
Major long-term projections of residential wood fuel usage found in
the literature included the following:
Office of Technology Assessment, Energy from Biological Processes,
I9609
U.S.F.S., RPA,.An Assessment of the Forest and Rangeland Situation in
the U.S., 1980lu ~~
Solar Energy Research Institute, Report on Building a Sustainable
Energy Future, 198111
32
-------
Booz, Allen, and Hamilton, Assessment of Proposed Federal Tax Credits
for Residential Wood Burning Equipment, 1979^
Bradburd, Mead, Schneider, and Art of Williams College, The Use of
Wood for Fuel; Historical Series and Projections to the Year 2000,
197913
14
Bonneville Power Administration (Steve Onisko)
Marshall, Thayer School of Engineering, Dartmouth College, The
Dynamics of Residential Wood Energy Usage in New England, 1970-2000,
T98I1"5
The first five studies above were national in scope and therefore not
suitable as a basis for regional projections. Their methodologies are dis-
cussed below. Residential wood combustion is a localized phenomena because
of high transportation costs, making national trend data of little or no
value for specific locations. The last two were regional projections, but the
methodology employed by Marshall was clearly superior to that employed by
any of the other studies. Only Marshall's report included a review of other
projections studies and the assumptions and causal factors behind other
researcher's projections. Each of the above projection study methodologies
is briefly described below.
9
1. OTA Projection
The Office of Technology Assessment's (OTA) report, Energy from Biological
Processes, makes explicit projection of residential wood fuel use (1.0 quads
if "business as usual" and 2.0 quads with "vigorous support and high energy
prices"), but there is no firm analytical basis for the projection. These
figures are based on Booz, Allen, and Hamilton estimates for 1985 with soft
assumptions that growth will occur if low-cost firewood remains available and
wood users continue to tolerate wood's lesser convenience than conventional
fuels. OTA's soft basis of estimate and the fact that it is a national projec-
tion make this method unsuitable for projections in the Pacific Northwest.
33
-------
2. U.S.F.S. Projections10
The U.S.F.S. projects that residential wood fuel use will increase from
. six million cords annually in 1976 to 26 million cords in the year 2030.
Based on these values, a year 2000 estimate of about 15 million cords is
derived with linear interpolation and a 12 million cord estimate is derived
with exponentialcurve interpolation. Projections are a function of past use
L trends, population trends, fuel price trends, and fireplace and stove sales
r trends. However, it is questionable whether it is valid to extrapolate such
short-term trends on a long-term basis._ The national nature of the projections
i also make these projections unsuitable as the single basis for projecting
Pacific Northwest wood usage.
k 3. SERI Projections11
P- The Solar Energy Research Center projections are presented in A New
Prosperity: Building a Sustainable Future. SERI projects annual residential
[ wood usage peaking in the 1980's in the 1.5 to 2.0 quads range and subsequently
dropping to about 1.0 quads per year in the year 2000. SERI's projections
are based on consumers' ultimate desire for convenience, improved heating .
efficiency, other forest land and wood energy use and historical wood use.
However, the calculation procedure and its basis are not explained in much
detail; and the national focus of the projection cannot be directly applied to
the Pacific Northwest.
4. Bradburd, et.al., Projections
Bradburd, et.al. projected residential wood use to the year 2000 in
The Use of Wood for Fuel - Historical Series and Projections to the Year 2000
under contract to the US DOE. The model is an optimization model under which
34
K.
-------
an individual is assumed to minimize his cost of space heating requirements,
wood price, conventional fuel prices, installation cost, and wood use per
household. The model assumes that these factors will basically remain the
same through the year 2000. Since these factors can be expected to change
significantly over the next 20 years, the model appears to have limited
validity for long-term projections.
12
5. Booz, Allen, and Hamilton, Inc. Projections
Booz, Allen, and Hamilton, Inc. projected future wood usage in Assessment
of Proposed Federal Tax Credits for Residential Wood Burning Equipment produced
under DOE contract. Oil displacement by wood usage is predicted and then
converted to estimates of wood usage. Wood usage of 0.8 quads is projected for
1985 (no 2000 projections) with 1.0 quads consumption rate projected with 30%
tax credit for purchase of wood heating equipment. The projections for future
consumption without a tax credit are based on historical wood burning equip-
ment sales and manufacturer's projections of sales. Extrapolation of such
short-term trend factors does not appear to be a valid method for 20 year
projections when the economic factors motivating increased wood usage in recent
years can so easily change over a 20 year period.
14
6. BPA Projection
14
In 1980, the Bonneville Power Administration (BPA) projected that by
the year 2000 RWC could offset 4.6% of the electrical space heating in the
Pacific Northwest, saving 3,720 million kwh of electricity. This assumed
that 40% of the estimated 1,943,700 electrically heated homes in the year 2000
would use wood as a primary or secondary heat source. This represents an
increase in wood fuel use of about 635,000 cords/year, between 1980 and 2000,
35
-------
throughout the Pacific Northwest. While interesting, these projections dealt
only with electrically heated households (estimated to account for 60% of
all households by 2000), so the approach was not comprehensive enough for
this study.
7. Marshall's Projections
Marshall projected residential wood usage in a thesis completed June,
1981, entitled The Dynamics of Residential Wood Energy Use in New England
1970 - 2000. Review of the work by Marshall indicated that it had the most
sophisticated and best documented methodology for projecting wood usage. It
included full documentation and listing of the computer model used, as well
as an explanation of which variables would need to be altered to apply the
model to other regions of the U.S. This model takes into account the most
variables expected to influence residential wood combustion levels of any of
the projections reviewed.
Marshall's work was funded by the U.S. Forest Service Forest Product's
Laboratory in Madison, Wisconsin and the Forest Economics subsection charged
with making national projections for residential .wood burning. Mr. Ken Scog,
U.S.F.S project monitor for Marshall's work, confirmed that the Marshall
report was the state-of-the-art methodology for projecting residential wood
usage within a specific region. He felt that applying the Marshall model to
the Pacific Northwest should be a most reasonable basis for projecting future
residential wood usage in the Pacific Northwest. Accordingly, Marshall's
economic model was selected as the methodology used to estimate long-term
fuel use for the Portland, Seattle and Spokane areas.
36
-------
B. DESCRIPTION OF SELECTED ANALYTICAL APPROACH
1. Description of Marshall's Model
The driving force behind the spectacular increase in residential wood
combustion throughout the U.S. in the last decade has been the spiralling cost
of home heating with conventional fuels such as oil, gas, and electricity.
For example, Seattle distillate oil prices increased from 21 cents per gallon
in 1970 to $1.08 per gallon in 1980. The price for 100 therms of natural gas
in Seattle increased from $15.60 to $61.43 during that time period. The
increase in annual home heating costs from $100 to $150 to costs in the
vicinity of $500 to $1000 per year caused many homeowners to seek less
expensive ways of home heating. Many chose to install wood burning systems
for home heating.
Marshall's model for projecting wood usage is an economic model in which
percentages of the population are assumed to install wood heating systems based
on cost differentials between heating with wood and heating with conventional
fuels. A weighted conventional fuel heating cost (in proportion to usage of
oil, gas and electricity) is compared with the cost of heating with wood. The
magnitude of the potential fuel cost savings drives the installation of new
wood burning devices and determines the assumed rate of capacity utilization
(percent of total possible usage) of wood heating system.
Important non-economic factors incorporated in the model include 1) the
effect that self-cut wood has on people's perception of the wood cost; and 2)
the effect that inconvenience has on people's perception of wood heating costs.
The model also incorporates relationships that can be expected to change over
time, such as 1) the effect that market penetration will have on increasing
37
-------
wood heating system installation costs; and 2) market penetration effects on
increasing perceived inconvenience costs of wood heating. These effects are
consistent with common sense since-first installers will be people with easiest
installations (i.e., existing fireplaces) and least inconvenience costs (easy
access to wood, etc.).
Perhaps the most important feature of this model is the use of best
available information on the cost of heating with wood versus alternative
heating fuels (oil, gas, electricity) to simulate how these cost factors will
interact to influence future levels of &WC. All of the model's assumed
relationships are spelled out in understandable form and well documented
enabling future researchers to build on a solid framework rather than having
to start from scratch. New changes can be easily incorporated into the model .
Marshall specifies the assumptions incorporated in the model which are listed
verbatim in Figure 4.
The model also includes a coal module and a pollution sector (where
particulate emission problems can be assumed to lead to regulations, which
increase the cost of new wood burning units). Both the coal and pollution
modules are set to have neutral effect in Marshall's New England projections
and our Pacific Northwest projections. The structure to incorporate these
effects has been included in the model, but does not affect the results. Thus,
researchers can easily experiment with the potential influence of coal usage
or pollution regulation if they so desire.
2. Addition of a Fireplace Wood Usage Sector
Although Marshall's work represents the best methodology found in the
1iterature-for projecting residential wood usage, it is deficient in that it
38 '
-------
Figure 4 Basic Assumptions In Marshall's Wood Use Projection Model
o Wood-heating equipment may be installed in new housing units or
retrofitted into existing housing.
o Decisions to invest in wood-burning equipment are made on the basis
of marginal costs and benefits.
o The decision to install wood-heating equipment in new housing is a
function of fuel savings and non-economic factors.
o The decision to retrofit wood-heating equipment into existing housing
is a function of economic payback period and non-economic factors.
o The economic payback period is equal to installation cost divided by
annual fuel savings.
o Installation cost reflects the amount of the housing with installed
capacity, the fraction of the household's heating needs met by the
wood-heating capacity, and the level of pollution abatement.
o The least expensive installations will in general be performed first.
Therefore, as the market is penetrated, installation cost increases.
o Early adopters with their own wood supply will have a lower perceived
wood-heating cost than later adopters who must purchase all of their
wood.
o Non-economic factors incorporated into the model are convenience and
pollution.
o Consumers for whom convenience is not a major issue will in general
install v/ood-burning equipment first. Therefore, as the market is
penetrated, inconvenience cost increases.
o Fuel prices, the size of the housing stock, heating efficiencies,
and household heating requirements are all exogenous inputs to WOODSTOV-2.
o Stove usage may displace a greater quantity of fuel than if the same
quantity of heat were provided by a central furnace.
39
-------
does not include projections for use of wood in fireplaces. This is because
Marshall and the sponsor, the USFS, were originally interested in the use of
wood for heating purposes. Fireplace heating efficiency is considered to be
nominal ranging from negative values up to 10 or 15%. Marshall and the USFS
regarded fireplace use as primarily for aesthetic purposes and of lesser
importance compared to wood usage in stoves and furnaces. However, one of the
early wood usage surveys conducted for the Portland area, the Talbot Wong
study showed that well over half of the wood burned in the Portland metro-
politan area in the 1977-78 heating season was burned in fireplaces. Given
the magnitude of fireplace wood usage compared to wood usage in stoves in the
early years of the wood heating expansion, the addition of a fireplace wood
usage sector was a decided improvement to the model .
Fireplace usage (as compared to stove, insert, or furnace usage) can be
expected to drop over time for a variety of reasons:
Fireplaces are very inefficient (from negative to about 10%
efficiency) when compared to wood stoves (40 to 60% efficient).
This explains why many households are willing to substitute stove
usage by installing them in existing fireplaces.
As fireplace installations are converted to stove installations,
wood usage in fireplace drops.
Increasing wood prices can be expected to reduce fireplace wood
usage. Aesthetic wood burning does not have the economic payback
achievable with stove wood burning.
Thus, the fireplace sector should utilize factors or relationships that
reduce fireplace wood usage over time. A review of several surveys, such as
the New England Fuel Wood Survey and Oregon wood heating surveys, confirmed
that this expected behavior did in fact occur in the last few years of the
1970 to 1980 time period. These survey findings are explained in detail in
Appendix A, Section Ic.
40
-------
Based on a review of alternate methods by which the expected fireplace
wood usage reduction could be "driven" (also explained in Appendix A), a
methodology was selected by which future fireplace usage drops as real wood
price increases in future years.
3. Limitations of Marshall's Model
Marshall's model does not directly take into account three major factors
that can influence wood burning levels. These factors are: the availability
of wood resources; the effect of various conservation efforts; and the effects
of various governmental policies to either encourage or discourage residential
wood combustion. All three factors will in turn influence other variables which
are included in the model, such as wood fuel price and household heating require-
ments. Further discussion of these three factors is included in Appendix B.
4. Calibration of the Model
a. Portland Model Calibration
A review of wood usage survey data for the three metropolitan areas
showed that no reliable estimates to confirm the model accuracy (calibrate
the model), were available for either the Spokane or Seattle areas. Survey
data for both fireplaces and stoves was available for the Portland area from
the Talbot Wong survey for the 1977-78 heating season. The Portland data was
used as the primary basis for model calibration together with 1970 census data
for primary heating with wood for Oregon as a whole and with DEQ/Talbot-Wong
estimations for 1980 wood usage in the Portland metropolitan area. Additionally,
some estimated stove sales data was available for the state of Oregon from the
Wood Energy Association in Portland.
The available data for use in model calibration was as follows:
41
-------
1977 Portland metropolitan area fireplace wood usage 209,000 cords/yr
1977 Portland metropolitan area stove wood usage 81,000
1980 Portland metropolitan area stove wood usage 144,000
1980 Portland metropolitan area fireplace wood usage 215,000
1978 Oregon stove sales 25,000 stoves/yr
1978 Portland area stoves sales inferred to be 12,500
1979 Oregon stove sales 33,000
1979 Portland area stove sales inferred to be 16,500
1980 Oregon stove sales 30,000
1980 Portland area stove sales inferred to be 15,000
In the calibration work, wood usage was considered to have greatest
importance. Sales data served as a secondary reference. Also, since DEQ's
fireplace wood usage projections did not project a drop in fireplace wood
usage, the DEQ estimate of 1980 fireplace wood usage was not utilized in
calibration work (with increasing wood prices, use of fireplaces should have
decreased, not increased, since it is largely aesthetic).
A detailed explanation of model inputs and changes is presented in
Appendix A. Figure 5 shows the model's portrayal of wood stove and fireplace
usage from 1970 to 1982 and the correlation with available data on actual
wood usage.
A good bit of known historical behavior is predicted with wood usage in
stoves accelerating in 1976. The 1977 stove wood usage is plotted at 80,000
cords, which is close to the estimate of 81,000 cords/year in Figure 6. The
1980 stove wood usage is shown at 150,000 cords/year, close to the 1980 estimate
of 144,000 cords/year. Wood usage in fireplaces is shown to begin to drop in
1979.
Stove sales, together with wood usage in stoves is portrayed in Figure 6
42
-------
Cords Wood/Year
cr o>
o
tu n>
IQ
o
o
o
CD
o
o
o
V
o
o
o
CD
o
o
o
o
o
o
<*
o
CD
o
10
--4
o
o>
-5
VO
00
o
*
*
*
cz
in
-l
ft) OJ
o ^
CL
O)
O 3
rt> rt>
c-f
z: -i
o o
o -o
Q. O
f- in r+
0- OJ Q)
£»:
Cu
o
CL
fD
01 -i
3 fD
Q- Q>
00 -H
rt O
O r+
< O>
(D
O O
O O
CL CL
in in
OJ 01
(O id
ft) fD
-------
Cords Wood/Year
£ (X)
O <-f
o o
Q. <
CD
CZ
V)
fa t/i
IQ Cu
ID
CD
in
O
O
O
O
O
no
o
o
o
o
o
U)
o
o
o
o
o
o
o
o
o
o
(D
Ol
-i
o
o
CL
OL
n>
rt-
-5
O
T3
O
00,^
£<£ -
CD
c:
o
I
O3
QJ
l/i
rt
O
Ou
rt>
-------
for years 1970 through 1982. Sales portrayed by the model are in the 13,000
to 20,000 per year range for the years 1976 through 1980 which is close to the
available estimates for the Portland metropolitan area in years 1978 through
1980. The model portrays a drop in stove sales the last part of the 1970's,
which is close to the year when sales actually did drop (1980). The fact that
stove sales do not exactly follow the estimated sales data is not of concern
since 1) Marshall states that the model should not be calibrated on sales, but
rather on wood usage; 2) the sales data is for Oregon only, rather than
Portland; and 3) the Wood Heating Alliance acknowledges that their stove sales
data has significant uncertainty (an estimate was made of how much of total
Oregon stove sales to all states were sales returning in Oregon installations).
Based on this close simulation of historical behavior, the Portland
metropolitan area model was considered to be calibrated.
Figure 7 shows the Portland metropolitan area BASE CASE projections for
residential wood usage through the year 2000 in stoves and fireplaces as well
as total wood usage. This is based on the most likely fuel cost escalation
factors and a 2% annual real price increase for cord/wood used by Marshall for
New England. Usage of wood in stoves is projected to more than double between
1980 and 2000, increasing from 150,000 cords per year to about 320,000 cords
per year in the year 2000. However, wood usage in fireplaces is projected to
drop from 200,000 cords per year in 1981 to 140,000 cords/year during that same
time period. The net effect of these opposite trends is that total wood usage
is projected to increase about 41% between 1980 and 2000 (340,000 cords per
year to about 480,000 cords per year in the year 2000).
The short-term trend analysis from recent years showed that residential
45
-------
FIGURE 7
Base Case Projection for the Portland Metropolitan Area
for
Stove and Furnace, Fireplace, and Total Wood Usage
(1970 to 2000)
500,000
400,000
ra
O)
-a
o
o
o
o
300,000
200,000
100,000
1970
Stove
Fireplace
T^ Total Usage
f-
1980
1990
20C
Year
46
-------
wood usage in Portland has been increasing at about 6.5% per year. The model
projections coincide almost exactly with this 6.5% growth rate. The three
numbers on Figure 7 at years 1981, 1982, and 1983 show expected total wood
usage would be with a 6.5% growth rate from a 1980 base of 340,000 cords
(373,000 short-term trend vs 380,000 modeled; 397,000 vs. 400,000 modeled; and
»
423,000 vs. 420,000 modeled).
Short-term extrapolations inevitably diverge from true behavior in any
systems with the passage of time, and the model projects a leveling in total
wood consumption beginning in the mid-1980's. For the period between 1984 and
1988, the model projects a 0% growth rate in residential wood consumption,
resuming a moderate growth rate averaging 1.6% per year from 1990 through the
year 2000.
Between 1980 and 2000, total wood usage is projected to increase only by
about 41%, from 340,000 cords/yr to about 480,000 cords/yr. However, the
air quality implications are more significant since stove wood burning is
expected to increase by about 125% (340,000 versus 150,000).
Most stoves are the airtight kind, which have more emissions per cord
than fireplaces. If DEQ-estimated emission factors were to remain constant
through the year 2000, total emissions from wood burning would increase by
72% as shown in Tables- This is discussed in more detail in the Portland area
conclusion section.
TABLE 4
Projected Range of Year 2000 Wood Usage by-Appliance
for the Portland Metropolitan Area
Appliance Range of Projected Wood Usage Best Estimate
Stove/Furnace 280,000 to 400,000 340,000
Fireplaces 100,000 to 150,000 140,000
Total Wood Usage 380,000 to 550,000 480,000
47
-------
Figure 8 shows the projection for wood usage in stoves and furnaces for
New. England as projected by Marshall. The general pattern projected shows the
same acceleration of use in the late 1970's, the same "slowdown" in the rate
of increase in the 1980's, and a similar reacceleration in the 1990's. Trends
in the 1970's in wood usage in both New England and the Pacific Northwest were
very similar. This can be ascertained by comparing the 1970 to 1980 portions
of the curves in Figures 7 and 8. The portions were calibrated on historical
use data. Thus, it is not too surprising that future year projections are
comparable in terms of the general shape of the curve.
b. Seattle and Spokane Model Calibrations
Similar calibrations of the model for use in Seattle and Spokane are
described in Appendix A (Sections 2d and 3d). The Base Case 1970-2000 pro-
jections which form the basis for the best estimates'of long-term trends in
wood usage (Section C) are also described in these appendix sections. Before
describing the Base Case best estimates, a preliminary sensitivity analysis of
the model is described in the following section again using Portland data for
illustration.
4. Sensitivity Analysis
Some sensitivity analysis was conducted for the Portland version of the
model in order to determine whether the model responds to changes in future
year conditions in a manner consistent with "common sense" expected results.
The primary factor increasing residential wood usage in this economic
model is the potential fuel cost savings, which is the difference between the
cost of heating with wood and the average cost of heating with conventional
fuels. Sensitivity analysis was performed to test the effect on projections of
48
-------
Cords Wood/Year
1
CD
0
0
CD
CD
rsj
0
0
0
o
o
co
0
0
CD
o
CD
^
0
0
o
o
CD
cn
O
CD
CD
0
0
UD
-^J
O
vo
co-
o
ro
o>
-t
UD
VD
O
O - _
o
o
O
o
Q-
C.
I/)
£D
IQ
UD 3
CD CO
(-)- O
O <
ft)
ro 1/1
CD
O £D
o a
ex
DJ
O
CD
in
cu
-s
o
ro
o
en
00
-------
either increasing or decreasing the cost of wood fuel relative to conventional
heating fuels. The projected real price growth rate for wood between 1980 and
2000 was 2% in the New England model base case. A similar rate was assumed in
the Portland base case. In the sensitivity runs described below, the effect on
projections was evaluated for two cases: 1) a constant real wood price (in
1980 dollars) between 1980 and 2000 instead of the 2% growth in real price used
in the base case; and 2) wood prices increasing at 5% (real growth) per year.
In both cases the real price escalation rates of conventional fuels (oil, gas,
and electricity) were held constant and the differences in projected wood fuel
use between 1980 and 2000 were noted.
a. Constant Wood Prices 1980-2000
Figure 9 shows the model projections for the Portland metropolitan
area if wood prices remain constant (in 1980 dollars) between 1980 and 2"000.
With no escalation in real wood prices, the usage of wood for heating in stoves
and furnaces increases to a level of 430,000 cords per year in 2000 as compared
to the base case projection of 340,000 cords per year for the year 2000
(Figure 8) or a 26% increase in usage above the base case projections. Thus
stove wood usage would be projected to increase from 150,000 to 430,000 cords/
year (187%) versus the base case projection of an increase from 150,000 cords to
340,000 cords (127%). As would be expected, fireplace wood usage is projected
to remain about constant under the wood price/fireplace usage assumptions input
into the model.
b. Five Percent Per Year Wood Price Escalation 1980-2000
On the other side of the scale, if real wood prices are projected to
increase at 5% per year, then wood usage is projected to peak in the mid 1980's
50
-------
FIGURE 9
Portland Metropolitan Area Wood Use Projections
with
Constant Wood Price from 1980-2000
(1970- 2000)
O)
T3
O
o
I/)
-o
o
o
600,000
500,000
400,000
300,000
200,000
100,000
1970
**
-A- -A-
1980
1990
2000
Stove
Fireplace
*Total Usage
Year
51
-------
and drop in the following years. This projection is shown in Figure 10.
Based on these sensitivity analysis .runs and others not discussed here,
it appears that the model does respond in a reasonable manner to input changes
that alter the potential savings achieved by converting to wood heating. For
example, with a 0% wood price escalation between 1980 and 2000, total year 2000
wood usage of 630,000 cords would be projected. The 5% escalation factor
results in usage of only 240,000 cords.
C. FINDINGS AND CONCLUSIONS
This analysis adopted and applied the state-of-the-art model for
projecting wood usage. Historical wood usage between 1970 and 1980 is well
predicted by the model. The model's predictions for 1981 through 1983 are
consistent with the short-term trend findings from Section II. This application
of the model added a fireplace usage sector, which was deemed necessary because
fireplace wood usage was over twice as great as wood usage in stoves in 1977
in Portland. Many important non-economic factors are incorporated in the model ,
and all assumptions are clearly described. The model is fully documented in
a manner that will allow interested researchers to input alternate assumptions,
factors, and price projections as better information becomes available. This
work was achievable within the scope of this contract because of Norman Marshall's
development of the model and complete documentation of its structure and input
assumptions. As in all projections, the accuracy of the final results is depen-
dent on the accuracy of the input data. Major deviations in future fuel prices
or availability from predicted values will likely result in substantial shifts
in levels of residential wood combustion from those predicted.
52
-------
Cords Wood/Year
I -n oo
O - rt
rt -S O
01 n> <
'X3 n>
cu
n
n>
o
o
<+
O
0
o
ro
o
o
<+
O
0
o
co
o
o
**
0
0
o
o
o
<*
O
0
o
CO
o
en
OJ
ft)
B»
10
o
ro
o
o
o
"O
-5
-n 73
-5 n>
o cu
TP
o
-j
Cu
3
CD-
CD
rt
-J
O
CU
UD O -
CO O rt
O CX 3-
cn
PD
rn
o -
o n
o n>
m
in
o
cu
cu
rt
O
o
ex
ro
"O
-s
o
t_l.
n>
n
rt
o
3
to
-------
1. Portland Metropolitan Area
The authors' best estimates of year 2000 total wood usage is about
480,000 cords/year. This represents the projections using the Bonneville
Power Administration's most likely price escalation rates for oil, gas, and
electricity and a 2% real price escalation factor for wood. Table 4 presents
year 2000 ranges of projected wood usage. Table 5 shows the best estimate
projections for years 1985, 1990, 1995, and 2000 for all three categories of
wood usage.
TABLE. 5
Best Estimate Projection of Residential Wood Usage
in
Portland for 1985 - 2000
Year Stove/Furnace Fireplace Total Wood
Wood Usage Wood Usage Usage
1980 150,000 190,000 340,000
1985 240,000 190,000 430,000
1990 240,000 170,000 410,000
1995 300,000 150,000 450,000
2000 340,000 140,000 480,000
Total wood usage is projected to increase at an annual rate of 6.5%
between 1980 and 1983 with a 0% growth rate projected between 1984 and 1988 and
a 1.6% growth rate between 1990 and the year 2000. A 41% increase in total wood
consumption is projected from 1980 to 2000.
The net growth rates projected above indicate a shift in the common devices
in which wood is projected to be burned. Stove wood usage is projected to
increase by 127% between 1980 and 2000, whereas fireplace wood usage is projected
to decrease by 26%.
54
-------
Even though total wood usage is projected to increase by only 41% between
1980 and the year 2000, emissions can be expected to increase more dramatically.
Wood burning in stoves produces more emissions per cord than fireplaces do and
is projected to increase by 127% in the 1980 to 2000 period. Table 6 shows
how emissions from the various wood burning categories can be expected to change,
using current DEQ emission factors for stoves and fireplaces. Even though the
projected increase in total wood burning in only 41%, the projected particulate
emissions increase (if emission factors were to remain constant) is 71%. This
clearly points out the importance and need for lower-emitting residential wood
burning stoves and fireplaces.
TABLE 6
Projected Change in Emissions
From
Portland Residential Wood Burning
Stoves Fireplaces Total
Year
1980
1985
1990
1995
2000
Wood
Burned
(Cd/yr)
150,000
240,000
240,000
300,000
340,000
TSP
Emissions*
(T/yr)
5,925
9,480
9,480
11,850
13,430
Wood
Burned
(Cd/yr)
190,000
190,000
170,000
150,000
140,000
TSP
Emissions*
(T/yr)
3,325
3,325
2,975
2,625
2,450
Wood
Burned
(Cd/yr)
340,000
430.JOOO
410,000
450,000
480,000
TSP
Emissions*
(T/yr)
9,250
12,805
12,455
14,475
15,880
* Assuming 1.75 tons wood/cord; fireplace and stove TSP emission factors of
20 and 45 "Ib/ton^O, respectively. Emission factors are assumed to remain
constant. Improvements in stove efficiencies will reduce emission rates.
55
-------
[
r
r_
2. City of Seattle
These authors' best estimates of year 2000 total wood usage in Seattle
city is about 185,000 cords per year. This represents the projections using
BPA's most likely price escalation rates for oil, gas, and electricity, and a
2% real price escalation factor for wood. Table 7 presents year 2000 ranges -
of projected wood usage and best estimates for stoves/furnaces, fireplaces, and
total wood usage. Table 8 shows the best estimate projections for years 1985,
1990, 1995, and 2000 for all three categories of wood usage.
TABLE_7
Projected Range of Year 2000 Wood Usage by Appliance
For the City of Seattle
(cords/year)
Appliance Best Estimate
Stoves/Furnaces 85,000
Fireplaces 100,000
Total Wood Usage 185,000
TABLE 8
Best Estimate Projection of Residential Wood Usage 1985-2000 in Seattle
(cords/year)
-Year Stove/Furnace Fireplace Total Wood Usage
j£ Wood Usage Wood Usage
1980 45,000 110,000 155,000
p 1985 85,000 100,000 185,000
I 1990 85,000 100,000 185,000
1995 90,000 100,000 190,000
r 2000 85,000 100,000 185,000
I 56
-
r
-------
Year 2000 total wood usage is projected to be 185,000 cords per year as
compared to 155,000 cords/year in 1980 for a net growth of about 19%. However,
wood burning in stoves and furnaces is projected to increase by 89% between
1980 and 2000, and fireplace wood usage is projected to drop by about 9%.
If ODEQ emission factors for stoves and fireplaces are assumed and no
improvements in clean burning technology occur, year 2000 TSP emissions will
increase by about 38% as compared to 1980, and almost all of this growth will
occur in the next five years.
Table 9 shows the projections for wood usage and resultant emissions
between 1980 and 2000 in five year intervals with assumed constant emission
factors.
TABLE 9
Projected Change in Emissions
From
Seattle City Residential Wood Burning
Stove / Furnace Fireplace Total
Year
1980
1985
1990
1995
2000
Wood
Burned
CCd/yr)
45,000
85,000
85,000
90,000
85,000
TSP
Emissions*
(T/yr)
1,770
3,350
3,350
3,540
3,350
Wood
Burned
(Cd/yr)
110,000
100,000
100,000
100,000
100,000
TSP
Emissions*
(T/yr)
1,925
1,750
1,750
1,750
1,750
Wood
Burned
(Cd/yr)
155,000
185,000
185,000
190,000
185,000
TSP
Emissions*
(T/yr)
3,695
5,100
5,100
5,290
5,100
* Assuming 1.75 tons wood/cord; fireplace and stove particulate emissions factors
of 20 and 45 Ib TSP/ton , respectively. Emission factors are assumed to remain
constant. Improvements in stove efficiencies will reduce emission rates.
57
-------
3. City of Spokane
The authors' best estimates of year 2000 total wood usage in Spokane city
is about 129,000 cords per year. This represents the projections using BPA's
most likely price escalation rates for oil, gas, and electricity, and a 2%
real price escalation factor for wood. Table 10 presents year 2000 ranges of
projected wood usage and best estimates for stoves/furnaces, fireplaces, and
total wood usage. Table 11 shows the best estimate projections for years 1985,
1990, 1995, and 2000 for all three categories of wood usage.
TABLE-10
Projected Range of Year 2000 Wood Usage by Appliance
For the City of Spokane
(cords/year)
Appliance Best Estimate
Stove/Furnace 54,000
Fireplace 75,000
Total Wood Usage 129,000
TABLE 11
Best Estimate Projection of Residential Wood Usage
for
The City of Spokane 1985-2000
Year Stove/Furnace Fireplace Total Wood Usage
Wood Usage Wood Usage
1980 28,000 93,000 121,000
1985 42,000 84,000 126,000
1990 45,000 si 000 126,000
1995 51,000 . 78',000 129,000
2000 54,000 75 000 129,000
58
-------
Year 2000 total wood usage is projected to be 129,000 cords per year as
compared to 121,000 cords/year in 1980 for a net increase of about 7%.
However, wood burning in stoves and furnaces is projected to increase by about
93% between 1980 and 2000, and fireplace wood usage is projected to drop by
about 19%.
If ODEQ emission factors for stoves and fireplaces are assumed and no
improvements in clean burning technology occur, year 2000 TSP emissions will
increase by about 26% as compared to 1980, and almost all of this growth will
occur in the next five years. _
Table 12 shows the projections for wood usage and resultant emissions
between 1980 and 2000 in five year intervals with assumed constant emission
factors.
A 14% increase in total emissions is projected between 1980 and 1985 even
though the total amount of wood burned stays about the same. This is because
airtight stoves are currently estimated to produce particulate emissions at
more than twice the rate as in fireplace combustion.
. TABLE 12
Projected Changes in Emissions from Spokane City Residential Wood Burning
Stove / Fnrnarp
Total
Year
1980
1985
1990
1995
2000
Wood
Burned
(Cd/yr)
28,000
42,000
45,000
51,000
54,000
TSP
Emissions*
(T/yr)
1,100
1,650
1,770
2,010
2,130
Wood
Burned
(Cd/yr)
93,000
84,000
81,000
78,000
75,000
TSP
Emissions*
(T/yr)
1,630
1,470
1,420
1,360
1,310
Wood
Burned
(Cd/yr)
121,000
126,000
126.,000
129,000
129,000
TSP
Emissions*
(T/yr)
2,730
3,120
3,190
3,370
3,440
Assuming 1.75 tons wood/cord; fireplace and stove TSP emission factors of
20 and 45 Ib/ton , respectively. Emission factors are assumed to remain
constant. Improvements in stove efficiencies will reduce emission rates.
59
-------
IV. REFERENCES
la. Memo report by Talbot, Wong and Associates, Inc., Consulting Engineers,
7 SE 97th*St., Portland, Oregon to the Oregon Department of Environmental
Quality (DEQ), 522 SW Fifth Ave., Portland dated Dec. 20, 1979, about
p estimated levels of residential wood combustion based on telephone surveys
| in Portland-Vancouver, Eugene-Springfield, and Medford-Ashland.
r Ib. Computer printout of telephone survey results by North Opinion Research,
I Inc., 1030 SW 13th Ave., Portland, Oregon 97205
[
[
I
Ic. DEQ internal memos or calculations "by R.L. Gay (Eugene-Springfield) and
Peter Bosserman (Portland-Vancouver), utilizing Talbot-Wong survey results
2. Personal communications from:
a. Don Carl ton, Mt. Hood National Forest
b. Jerry Hazen, Mt. Baker/Snoqualmie National Forest
c. Mike Griggs, Washington Dept. of Natural Resources, Enumclaw , WA
d. Tim Crotts, Weyerhaeusei; Inc., Enumclaw.. WA
e. Morey Vogle, USFS Information Office, Spokane, WA
f. Gene Hollater; Darrell Evens, Colville National Forest
g. Wayne Orr, Idaho Panhandle National Forest
h. Phil Hildebrand, Washington Dept. of Natural Resources, Spokane, WA
i. Mike Sullivan, Industrial Forestry Association, Portland, OR
j. Tim O'Donnell, NW Pine Association, Spokane, WA
k. Robert Dick, Washington Forest Protection Association
3. Portland State Implementation Plan (1980)
4. Personal communications from:
a. Mike Ogan, Metropolitan Service District (METRO), Portland, OR
b. Chandler Felt, King County Planning Department
c. Tim Watterson, Puget Sound Council of Governments
d. Doug Adams, Spokane County Planning Commission
60
-------
5. U.S. Department of Commerce, 1970 Census of Housing, Vol. 1, Part 49
(Washington), p. 49-81
6. U.S. Department of Commerce
7. Puget Sound Air Pollution Control Authority Air Quality Records
8. Oregon Department of Environmental Quality Air Quality Records
9. Office of Technology Assessment, Energy from Biological Processes, 1980
10. U.S.F.S., RPA, An Assessment of the Forest and Rangeland Situation in the
U.S., 1980
11. Solar Energy Research Institute, Report on Building a Sustainable Energy
Future. 1981
12. Booz, Allen, and Hamilton, Assessment of Proposed Federal Tax Credits for
Residential Wood Burning Equipment, 1979
13. Bradburd, Mead, Schneider, and Art of Williams College, The Use of Wood for
Fuel; Historical Series and Projections to the Year 2000, 1979
14. S.A. Onisko, "Biomass End-Use Resource Report", Division of Power Management,
Branch of Conservation, Bonneville Power Administration, October 30, 1980
15. N. Marshall, Thayer School of Engineering, Dartmouth College, The Dynamics
of Residential Wood Energy Usage in New England, 1970-2000, 1981 R.P.I363
16. Seattle City Light, Residential Customer Characteristics Survey, June 1981,
Seattle, Washington
17. Del Green Associates, Inc., Residential Wood Combustion in the Pacific
'Northwest, Task 2A, 1982
18. Puget Sound Power and Light, Residential Wood Heating Survey, July 1980
Seattle, Washington
19. Washington Water Power, WWP Residential Survey, November 1981
20. November 1981 communication with Barbara Tombleson, Oregon Department of
Environmental Quality
21. "Firewood Theft Survey - 1981", administered by the Washington Forest
Protection Association, 711 Capital Way, Evergreen Plaza Bldg., Suite 608,
Olympia, WA 98501
22. John A. Cooper, "Environmental Impact of Residential Wood Combustion
Emissions and Its Implications", APCA Journal, 30(8), 855-861, August 1980
61
-------
23. Personal communication from Mr. Robert Marsk'e, Industry Division, Bureau
of the Census, U.S. Department of Commerce, Washington, D.C. 20233
24. J.A. Cooper, J.G. Watson, and J.J. Huntzicker, "Portland Aerosol Character-
ization Study (PACS)", report to the Oregon Department of Environmental
Quality, April 23, 1979
25. J.A. Cooper, R. DeCezar, et.al., "Medford Aerosol Characterization Study
(MACS)", report to the Oregon Department of Environmental Quality
26. J.A. Cooper, L.A. Currie, and G.A. Klonda, "Application of Carbon-14
Measurements to Impact Assessment of Contemporary Carbon Sources on Urban
Air Quality"
27. J.A. Cooper, L.A. Currie, and G.A. Klonda, "Evaluation of Carbon-14 as a
Unique Tracer to Determine the Maximum Impact of Contemporary Carbon
Sources of Atmospheric Particulates_ on the Portland and Eugene Airsheds",
final report to the Oregon Department of Environmental Quality by the
Oregon Graduate Center, July 25, 1979
28. T.G. Esvelt and M.L. Roberts, "The Use of Wood for Residential Space
Heating in the Pacific Northwest", Bonneville Power Administration, PO
Box 3621, Portland, Oregon 97208, presented at Solwest 1980 Joint Solar
Conference, August 7, 1980, Vancouver, B.C. Canada
29. Personal communication with Dan Howe, Northwest Natural Gas, Portland,
Oregon
30. Personal communication with Mary Beth Corrigan, Oregon Department of
Energy
31. Personal communication with Jim Ranfone, Gas Appliance Manufacturers
Association, Arlington, VA
32. Personal communication with Glenn Harding, Oil Heat Institute of Oregon,
Portland, Oregon
33. Personal Communication with John Fontain, U.S. Bureau of Labor Statistics,
San Francisco, CA
34. Personal communication with Chris Barnes, Portland General Electric,
Portland, Oregon
35. Personal communication with Edna Page, Pacific Power and Light, Portland,
Oregon
36. Oregonian, classified section, December 1-15 of years 1970-1980, Portland,
Oregon
62
-------
37. Oregon DOE, Oregon's Energy Future, Fourth Annual Report, January 1, 1980,
Table A-14, p. 100
38. Oregon DOE, Oregon's Energy Future, p. 57
39. BPA, Pacific Northwest Electric Energy Model, outputs dated 7/30/81 and
most recent available in November, 1981
40. M. Bailey and P. Wheeling, "New England Fuelwood Survey", Economics,
Statistics and Cooperatives Service, U.S. Department of Agriculture,
1974, Sproul Road (4th Floor), Broomall , PA 19008, March 20, 1980,
Press Conference
41. Charles E. Hewett, "Institutional Constraints on the Expanded Use of Wood
Energy Systems", Paper presented at the Third Annual National Biomass
Systems Conference, at Solar Energy Research Institute, Golden, CO,
June 6, 1979, published by the Resaurce Policy Center (RP #155), Thayer
School of Engineering, Dartmouth College, Hanover, NH 03755
42. Washington Department of Natural Resources, Wood Waste for Energy Study,
Executive Summary, January 1, 1979, p. 11
43. Oregon Department of Forestry, Forestry in Oregon ... 1980 Oregon Timber
Supply Assessment, December 1980, p. 80
44. Dr. F. Bryan Clark, USFS, Energy from Forest Biomass, USFS, presented at
the Interagency Workshop on Biomass, May 14-15, 1981, Washington, D.C.
p. 8
45. U.S. Forest Service, An Analysis of the Timber Situation in the United
States 1953 -2030, 1981
46. Personal communication with Bill Day, Anchor Tools and Wood Stov.es,
January 1982
47. N.E. Fuller, P.E., Bonneville Power Administration, Forest and Mill Residue
Resource Assessment, March 1980
48. See Appendix A.I.(4) for documentation
49. See Appendix A.I for documentation
50. Oregon DOE, 1980 Oregon Timber Assessment, p. 34
51. US. Forest Service FSM 2/80 Amend 42, Part 223, Sale and Disposal of
Timber, Part 223.1
63
-------
52. John J. Garland, Forest Engineering Department, Oregon State University,
Commercial Wood Fuel Harvesting, 1981, citing David M. Smith, Yale
University, in "Green America", American Forest Institute, Washington, D.C.
1979
53. Grants Pass Daily Courier, editoral entitled "The Need for Firewood",
August 31, 1979, reprinted from the Salem Capital Journal
54. Weatherization reduction by 80%.
55. , Portland Energy Office, February 1982
56. Oregon Department of Energy, Weatherization, One Step At A Time, 1980
57. Civil Engineering - ASCE, "Zero-Energy House: Bold, Low-Cost Breakthrough
that May Revolutionize Housing", May 1980, p. 48
58. Residential Solid Fuels, Environmental Impacts and Solutions, Proceedings
of the 1981 International Conference on this topic, held June 1-4, 1981 in
Portland, Oregon, Edited by John A. Cooper, sponsored by the Oregon
Graduate Center, Beaverton, Oregon. Papers included:
a. W.T. Greene and B.J. Tombleson, "Institutional and Regulatory Approaches
to Control Residential Wood Burning Emissions", pp. 1229-52
59. Time Magazine, Economy and Business Section, February 1, 1982
64
-------
APPENDIX A
DETAILED DOCUMENTATION OF PROJECTION MODEL CALIBRATION
-------
APPENDIX A
DETAILED DOCUMENTATION OF PROJECTION MODEL CALIBRATION
At a minimum, application of Marshall's model to a particular area or
region requires inputs of region specific information on local fuel prices,
local annual BTU heating requirements, local population, etc. Additionally,
some of the model's internal table functions defining relationships between key
variables within the model's program must be modified since they assume New
England type conditions, such as a 100 million BTU annual home heating
requirement (the Pacific Northwest requirement is about half of that). Lastly,
in the Pacific Northwest application of the model, a fireplace wood usage
sector was added that was not included in the original model. These three
categories of modifications are explained in detail; and model calibration
documentation is provided for Seattle and Spokane:
o local condition inputs
o modifications to table funct-ions
o addition of a fireplace wood usage sector
o model calibration (Seattle and Spokane only)
The application of the model to the Pacific Northwest could not have been
achieved without the superlative job of model documentation presented
throughout Norman Marshall's report. Rather than repeat his documentation in
detai'l, those readers interested in the specific logic behind the original
model are advised to obtain a copy of his report from the Resource Policy
Center at Dartmouth College in New Hampshire.*
*The Dynamics of Residential Wood-Energy Use in New England 1970-2000, June
1931 Masters Thesis by Norman Marshall.15 Available from the Resource Policy
Center as RP#363, Thayer School of Engineering, Dartmouth College, Hanover, NH
03755.
A-l
-------
i. Portland Model Documentation
A. L 3c11 Condition Incuts
(1) Initial Households Total
250 C THI = 341,500
jh r^mher of households in the Portland-Vancouver metropolitan area in
'' '"v ;! ^ns'-'iio ! ; Growth FMnet ion
'-;:0 A HGF ,< = TABLi fTHHF, TTIE.K, 1070, 2000, 5^
300 I THG." = .03G/.0307.0.75/.023/.0217.019/.019
A 'o.-;-!1 -)]?"niiq agency supoli-?d t'1? following dat a . ° a Th^ seven numbers
in t^a i~ol? in Line 300 represent annual household growth rates at five-year
intr-rv?'s between 1°70 percentaqos and 2000. Agency staff supplied household
esti-n^' of 341,505 for 1970, 471,850 for 1930 and 735,550 for 2000. The
staff recommended t^at straight linear interpolation should be assumed rather
t;',an an exoonsntial qrowth rate. T^ese estimates of households were linearlv
^nto-nn ,?.r-:-d = c five year intervals, drvj the growth rates in Line 300 represent
t'v° an;!1;,-;! nr^'-.'th rates which yield the above household values for 1930 and
200'I.
' 3 ^ Initial Fraction of Households vn'th Wood Heating
r/0 C IFHWC = .047
A 1970 estimate of the fraction of households heating with wood in Oregon
/as 'vaii^bip in a 1930 paper by Esveldt and Roberts?! of the Bonneville Pov/er
Administration. Lacking Portland data, this value was used as the baseline
ino'jt. As an aside, year 2000 projection levels are nor. affected whether this
initial input is varied between 1 and SY,.
(I) Heating Requirements Per Household
1040 A HRH.K = TABLE (THRU, TIME.K, 1970, 2000, 5)
1050 T THRU = 50E6/48E6/45E5/42.5E6/40E6/38.5E6/37.5E6
The values in Line 1050 represent that annual average household heating
requirements dropped from 50 million BTU's in 1970 to 45 million BTU's in 1980
and wi 11 drop to 37.5 million BTU's per year in 2000. These figures represent
a 25!* droo in average household heating requirements between 1970 and 2000
which is consistent with the New England projection data in Marshall's model.
The actual average heating requirement per household is difficult to know since
a varietv of estimates are available and depend on assurnotions about age of the
housing stock, type of fuel, efficiency, insulation levels, square footage
assumptions, etc.
Northwest Natural Gas estimates that 1000 therms were needed to heat the
1000 square foot house in 1972.?? The average home sold in the
A-2
-------
Portland area in 1980 had 1332 square feet, which suggests average gas
retirements in 1972 would have been 1272 therms (source: Real Estate Report
for Metropolitan Portland, Fall 1981. Clackamas = 1447 sq. ft.; Washington -
1413 sq. ft.; Multnomah = 1103 sq. ft.; Clark = 1355 sq. ft.; population
weighted average of 1272 sq. ft.; Portland Real Estate Report, Box 19836,
Portland, Oregon). At 40% (Oregon Department .of Energy)23 to 50% (Marshall,
and Gas Appliance Manufacturer's Association, Arlington, VA)
-------
(6) Historical Oil Prices
1160 A OP1.K=TA8HL(TOP1,TIME.K,1970,1980.1}
1170 T T0?l = .4097.424/.4227.4807.5817.597/.6347.6097.5857.8217.974
Historical distillate oil prices for the Portland area were obtained from
the Oil Heat Institute. Table 1& shows orices in actual dollars (rather than
real 1980 dollars) in Column 2 with prices in real 1980 dollars shown in Column
3.
TABLE 14
Historical Portland Area Oil Prices
in Actual and 1930 Do""ars
(Price/Gallon)
Year Actual Price (S) Price in 1980 (S)
1970 .135 .409
1971 .199 .424
1972 .204 .422
1973 .239 .480
1974 .313 .581
1975 .372 .597
1976 .420 .534
1977 .429 .609
1978 .449 .585
1979 .711 .821
1980 .974 .974
(7) Historical Gas Prices
1250 A GP1.K=TABHL(TGP1,TIME.K,1970,1930,1)
1250 T TGP1=2.82/3.22/3.61/3.85/4.01/4.09/4.65/5.13/5.46/5.53/5.43
Historical gas prices for the Portland area were obtained from Northwest
Natural Gas.22 1970 to 1980 prices in actual and 1980 dollars per therm are
shown in Table 15.
A-4
-------
TABLE 15
Historical Portland Area Gas Prices
In Actual and 1980 Dollars
(S/Therm = S1Q5 BTU's)
Year Price in 1980 ($)
1970 2.82
1971 3.22
197? 3.61
1973 3.85
1974 4.01
1975 4.09
1575 4.56
1977 5.13
197S 5.46
1979 _ 5.53
5.43
(8) Historical Electricity Prices
1340 A EP1.K=TA3HL{TEP1,TIME.K,1970,1980,1)
1350 T TE?1=.022127.02137.02077.02797.02377.03147.03437.03557.03357.03137.037
Historical electricity prices for the Portland area were obtained from
Portland General E1ectric27 and Pacific Power and Light.28 These values, in
S/'
-------
Since PPL and PGE in 1980 served 30.2% and 69.8% of the households in the
Portland area, the prices in Table 15 above were weighted in those proportions
to derive average prices in the region. Table 17 shows those weighted orices
in actual dollars in Column 2 and in 1980 dollars in Column 3.
TABLE 17
Portland Metropolitan Area Weighted Average
Electricity Costs in Actual and 1980 Dollars
(SAwh)
Year Actual Price 1980 Price
1970 .0100 .0221
1971 .0100 _ .0213
1972 .0100 .0207
1973 .0139 .0279
1974 .0157 .0287
1975 .0195 .0314
1976 .02274 .0343
1977 .0250 .0355
1978 .02575 .0335
1979 .02701 .0313
1980 .03761 .0376
(9) Historical Wood Prices
1440 A WP1.K=TABHL(TWP1,TIME.K,1970,1980,1)
1450 T TWPl=55/53/62/70/73/75/83/89/101/98/90
Wood prices per cord for each of the years 1970 through 1930 were
determined by a review of average prices advertised in the Classified Section
of the Oregonian between December 1st and 15th of each of those years. Table
18 shows those prices in actual cost in Column 2 and in 1980 cost (via CPI) in
Column 3.
A-6
-------
TABLE 13
Wood Prices 1970-1930 in Actual and 1930 Dollars
(S/cord)
Year Actual Dollars29 1980 Dollars
1970 25 55
1971 25 53
1°72 30 62
1973 35 70
1974 40 73
1975 47 75
1976 55 33
1977 53 39
1978 73 101
1979 85 98
1930 90 90
(10) Historical Conventional Furls Usage Split
1070 A FCFO.K=TABLE(TFCFO,TIME.K,1970,2000,5)
1080 T TFCFO=.4587.3917.3017.267.2357.2177.200
1090 A FCFG.K=TA5LE(TFCFG,TIME.K,1970,2000,5)
1100 T TFCFG=.2567.2627.2637.257.257.257.25
Since the model compares wood heating costs to average costs of heating
with conventional fuels, the costs of the different conventional fuels must be
weighted in proportion to households that heat with those various fuels.
However, this split information was only available for 1975 and 1930 for Oregon
as a whole. Since Bureau of Census data was available for the Portland
metropolitan area for 1970, the fuel splits for year 1975 and 1980 were derived
by projecting changes in Portland area fuel splits based on how the splits for
the whole state of Oregon changed between 1970 and 193Q. Table 19 below shows
the resulting estimated relative usage of conventional fuels.
TABLE 19
Portland Area Shares of Household Heating
Via Conventional Fuels, 1970-1980*
(Shares Relative to 1.00)
Year Oil Heating Gas Heating Electricity Heating
1970 .458 .256 .285
1975 .391 .252 .347
1980 .301 .253 .435
*The fraction of conventional heating from oil and gas are input
in the model in Lines 1080 and 1100, respectively. Electricity shares are
calculated in the model by difference. Splits for 1985, 1990, 1995, and 2000
are also input in these same line numbers as explained in the following.
A-7
-------
(11) Future Conventional Fuel Usage Split
Future projections for the splits between usage of conventional fuels v/as
only available for the state of Oregon as a whole.30 Therefore, the 1980 split
for the Portland area was adjusted for future year split projections based on
how the statewide splits are projected to change. Table 20 shows those future
year projections.
TABLE 20
Portland Area Projected Shares of Household Heating
Via Conventional Fuels, 1985-2000*
(Shares Relative to 1.00)
Year Oil Heating Gas Heating Electric Heating
1985 .26 .25*2 .483
1990 .235 .252 .513
1995 .217 .251 .532
2000 .200 .249 551
#0il and gas data are input in Lines 1030 and 1100, respectively. Electricity
share is calculated in the model by difference.
(12) Future Oil and Gas Prices
1180 A OP2.K=TABHL(TOP2,TIME.K,1980,2000,2.5)
1190 T TOP2=.974/1.09/1.213/1.30/1.388/1.512/1.636/1.705/1.773
1270 A GP2.K=TABHL(TGP2,TIME.K,1930,2000,2.5)
1280 T TGP2=5.43/6.30/7.17/7.82/8.53/9.30/10.13/10.94/11.81
ODOE has published projections for real rates of price increase for
residential distillate oil and natural gas.31 These price projections were
used to derive the oil and gas price projections shown in Table 21. Table
increments are 2.5 years to be consistent with the model's input requirements.
A-8
-------
TABLE 21
Projected Portland Area Future Residential
Oil and Gas Prices
(1930 dollars/gallon) (1930 dollars/therm)
Oil Price Gas Price
1930 .974 5.83
1932.5 1.09 6.30
19S5 ' 1.213 7.17
1937.5 1.30 7.82
1990 1.338 8.53
1992.5 1.512 9.30
1995 1.536 10.13
1997.5 1.705 10.94
2000 1.773 11.81
(13) Future Electricity Price
1350 A EP2.K=TABHL(TEP2,TIME.K,1980,2000,2.5)
1370 T TEP2=.0376/.03S7/.03997.04157.04327.05184/.06037.05446/.06902
The Sonneville Power -Administrati on (Division of Power Requirements,
Division of Conservation) has projected future electricity prices for private
utilities. Since BPA is the designated agency for regional energy planning,
the electricity price escalation factors from BPA's "Pacific Northwest Electric
Energy Model"32 (7/30/81 run, still the most recent version available on
11/23/81) were used to project future electricity prices. Projected real price
escalation factors are 80-85:1.0503, 85-90:1.0339, 90-95:1.4076, and
95-2000:1.0949. Electricity price projections are shown in Table 22 with table
entries 2.5 years apart to be consistent with model input format.
TABLE 22
Future Portland Area Electricity Price Projections
($/kwh)
Year Price in 1980 $
1980 .0376
1982.5 .0387
1985 .0399
1987.5 .0415
1990 .0432
1992.5 .05184
1995 .0608
1997.5 .06446
2000 .05902
A-9
-------
(14) Future Wood Prices
A W?2.K=TABHL(TWP2, TIME. K, 1980, 2030,?. 5)
1470 T TWP2=90/95/99/104/109/115/121/127/133
No official government agency price projections were available for
residential cord wood. Neither the Wood Heating Alliance, Wood Energy
Association, nor the U.S.F.S.'s Ken Scog were aware of any such projections.
Accordingly, the base case wood orice escalation factors from Marshall ' s New
England work (2%/year real price growth) were assumed to apply. Some
rpsearch?rs have noted wood applications that ai^ enemy conversation and have
r.:iqnest?d wood nrices will escalate as fast as r. invent i ona 1 energy prices.
During the 19SO to 2000 period, conventional fu3! nrices in the Portland area
aro Dro;-?ct2d to increase at an annual rat3 of ?.OS^. Thus, a wood price
es-ca 1 a t i on vate of 2?'/vear sesrns reasonahle, an'i w^s therefore utilized. Table
93 shows wood price projections for 1^30-POQj aaa'ii in 2.5 year increments to
f^e consistent with model input requirements.
TABLE 23
Portland Area Wood Price Projections 1930-2000
(S/corH)
Year Price
1980 90
1982.5 95
1935 99
1987.5 104
1990 109
1992.5 115
1995 . 121
1997.5 127
2000 133
(15) Fuel Conversion Efficiencies
1130 A OE.K=TABLE(TOE,TIME.K,1970,2000,5)
1140 T TOE=.5/.5/.55/.575/.6/.625/.55
1310 A EE.K=TABLE(TEE,TIME.K,1970,2000,5)
1320 T TEE=.95/.957.95/.957.95/.95/.95
1220 A GE.K=TABLE(TGE,TIME.K,1970,2000,5)
1230 T TGE=.6/.67.fi/.6257.557.6757.7
A-10
-------
TABLE 24
Efficiency of Heatinq with Conventional Fuels
(relative to 1.00)
Year Gas Oil ' Electricity
1970 .600 .500 .95
1975 .600 .500 .95
1980 .600 .55 .95
1935 .625 .575 .95
1990 .650 .600 .95
1995 .675 .625 .95
2QQO .700 .650 .95
B. Explanation of Table Functions Adjusted
In systems dynamics modeling, simulated relationships are often attempted
that difficult to define in the form of a y=x, etc. equation. In systems
dynamics work in the Dynamo programming language, in cases where the general
relationship between factors is thought to be understood, table functions are
often employed to quantify the relationship between variables. For example,
one variable named TEMPWIC, which stands for "The effect of market penetration
on wood installation cost," is used to represent the relationship that wood
installation cost increases as more of the market is penetrated (since the
easiest and cheapest installations are the first ones generally to install wood
stoves). The particular program lines which specify this relationship are:
450 A EMPWIC.K = TABLE (TEMPWIC,HWC.K/TH.K,0,1,2)
460 T TEMPWIC = 1/1.5/3/5/5/5
Without explaining all the details about Dynamo programming language, the
reader can note that the variables generally are a mnemonic device that
abbreviates the variables they represent. HWC and TH refer to "households with
wood caoacity" and "total households," respectively. The degree of market
penetration ranges from 0 to 1.00 and represents the fraction of houses with
wood capacity divided by total households. Table 25 shows in more conventional
format the relationship which is defined by Lines 450 and 460.
A-ll
-------
TA5LE 2^
Market Penetration Table
Market Penetration or Fraction of
'Stove) Installation Capacity Is Multiplier by Which Baseline Wood
Expected to Increase Households with Wood Capacity
0 1.0
.2 1.5
.4 3.0
.5 5.0
.8 5.0
1.0 5.0
Now that this introduction has been completed of how table functions are
used in the program, it is appropriate to explain which table functions v/ere
changed in the model and on what basis. Marshall's report The Dynamics of
Residential Wood Usaoe in New England 1970-2000 provides complete documentation
of why various table functions were used, etc. Readers interested in
understanding the model in detail can review that work. The following
discussion explains only those table functions that were modified for the
Portland area.
Marshall specifies which table functions that should be modified when
applyinq the model to other areas. Those variables (asterisked) and the others
modified are listed in Table 26 along with an explanation the variable.
TABLE 26
Variables Modified in Portland Model Runs
*TNWIC Table for Normal Wood installation _Cost
TCFI .Table for .Capacity fraction ^Installed
*TCIWC Jable of the Cost of the ^convenience of Wood from Capacity
*TESCWP Table for the Affect of .Self Cut Wood on Ptice
TFCIWC Table which defines the fraction (of households) from £ost
(i.e. potential savings with wood heat) which will J/istall
Wood Capacity
A-12
-------
(1) Normal Wood Installation Cost (TNWIC)
In the Mew England version of the model, the following tabular
relationship is employed.
Capacity Fraction Installed
0
.2
.4
.6
.8
1.0
Normal Hood Installation Cost
200
300
400
600
1000
1500
Since the annual BTU heating requirement For the Portland metropolitan
area is half of that in the New England model, the cost of installing 100% wood
heating capacity was assumed to be exactly half of the value for New England,
or $750. The cheapest stoves available in the Portland area are about $200.
Thus, these end-points were used for the* Portland area and intermediate values
for .2, .4, .5, and .8 capacity were derived by assuming similar ratio
interpolations for the intermediate points. The Portland area relationship is
shown in Table 27. Note the cost of installing 50% capacity is set to equal
S370, or $200 plus 400/1300 of the magnitude between the table end points
(750-200), analogous to how the $600 cost for 60% capacity in New England
equals 200 plus 400/1300 times the magnitude between table end points, which is
1500 minus 200, or 1300. This type of interpolation, rather than linear
interpolation is appropriate because the cost of increasing capacity installed
from 30% to 100% is hiqher than the differential cost of increasing installed
capacity from 20% to 40%.
TABLE 27
Portland Area Normal Wood Installation Cost
. Versus Capacity Fraction Installed
Capacity Fraction Installed
0
.2
.4
.5
.8
1.0
Normal Wood Installation Cost
200
240
285
370
540
750
The specific lines in the model which incorporate this relationship are
shown below.
470 A NWIC = TABLE (TNWIC, CF1.K, 0, 1, .2)
480 T TNWIC = 200/240/285/370/540/750
A-13
-------
(2) Capacity Fraction Installed (TCFI)
This table function represents how the capacity fraction installed by
households varies as the average fuel cost savings varies. Table 28 shows the
relationship in the New England version.
TABLE 28
New England Average Fuel Cost Savings
-200
0
200
400
600
800
1000
vs. Capacity Fraction Installed
.2
.3
.4
.5
.8
.9
.95
This table function was changed in the Portland model. With only a 50
million BTU heating requirement in the Portland area as compared to a 100
million BTU requirement in New England, fuel cost savings will necessarily be
less and a certain capacity fraction represents different sizes of heating
equiment. For example, 90% capacity fraction represents 90 million BTU's
heating capacity in New England, but only 45 million BTU's heating capacity in
the Portland area.
For the Portland area, the following table function was employed:
TABLE 29
Portland Area Average Fuel Cost Savings
vs. Capacity Fraction Installed
Average Fuel Cost Savings
-100
0
100
200
300
400
500
Capacity Fraction Installed
.5
.6
.7
.8
.9
.95
.95
The basis for these table values is as follows. Based on the New England
table, a -$100 savings is associated with installation of 25% capacity which is
a stove with 25 million BTU's capacity. For Portland, it was similarly assumed
a -S100 savings was associated with a 25 million BTU stove, but this represents
capacity. Also, in New England, a S200 savings i.s associated with
A-14
-------
capacity installation, which is 40 million BTU's. In Portland, a $200 savings
is associated with a 40 million BTU caoacity stove, but this represents 30%
capacity. The maximum capacity fraction installed values are set at 95% since
stoves commonly have trouble heating a house 100% (the systems for transferring
heated air are often less sophisticated than with conventional fuel heating
equipment). This paragraph explains rows 1, 4, and 6 and 7 of Table 24;
remaining table values were interpolated.
The model line numbers which incorporate these table values are:
520 A CF1.K = TABLE (TCFI, AFCS.K, -100, 500, 100)
530 T TCFI = .5/.6/.7/.S/.9/.95/.95
(3) Cost of the Inconvenience of Wood from Capacity (TCIWC)
The model incorporates a variable TCIWC which stands for the Cost of the
Inconvenience of Wood due to Capacity utilized representing the relationship
that wood heating inconvenience costs go up as a greater percentage of home
heating is achieved via wood. Table 30 shows the New England values assumed in
Column 2 and the Portland area values "assumed in Column 3. Portland values
were simply set at 50% of the New England values since the heating requirement,
and therefore the amount of wood heating and associated inconvenience, is only
50% for the Portland area compared to New England (50 vs. 100 million
BTU's/year).
TABLE 30
Relationship Between Capacity Fraction Installed
and the Cost of Inconvenience of Wood due to Capacity Utilized
Wood Heating Capacity Utilized New England CIWC Portland CIWC
0 200 100
.2 300 150
.4 500 250
.6 800 400
.8 1000 500
1-0 1000 500
The line numbers in the Portland model that represent these values are
shown below. Capacity utilized in Line 990 is the fraction AC/HRH or average
capacity/household heat requirement.
990 A CIWC.K = TABLE (TCIWC, AC.K, HRH.K, 0, 1, .2)
1000 T TCIWC = 100/150/250/400/500/500
(4) Effect of Self Cut Wood on Price (TESCWP)
The TESCWP variable stands for the Effect of Self Cut Wood on Price and
represents the relationship that people perceive their wood acquisition costs
are less when they can cut their own wood rather than having to purchase it.
A-15
-------
The amount of forest resources available per capita is considerably higher in
Oreqon than in the New England states. For example, the average amount of
"cords of growing timber stock per household" is 1104 in Oregon versus only SO
in New England.* Also the amount of "growing stock growth in cords per
household" is 15.7 in Oregon versus only 1.03 in New England.* Thus it was
assumed reasonable to use a different "self-cut wood effect on price" variable
for the Portland area which would result in perceived price being somewhat
lower in Oregon than in New England. Table 31 shows the table function values
assumed for New England in Column 2 and the Portland area in Column 3.
*;Xarshan, op.cit., citing the USDA Forest Service, Forest Statistics of the
U.S., 1977, Washington, D'.C., 1978, p. 1-2, 33-34, 97-98.
TABLE 31
The Effect of Self Cut Wood on Price
Versus Wood Heatfng Penetration
Market Penetration ESCWP New England ESCWP Portland
0 .5 .4
.2 .8 .5
.4 1.0 .8
.6 1.0 1.0
.8 1.0 1.0
1.0 1.0 1.0
One additional modification was employed. In the New England model
version, penetration was defined as the ratio of houses with wood capacity to
total households. However, since the 1980 to 2000 growth rate in households is
projected to be considerably higher in the Portland area (56% vs. 22%), this
more rapid household growth rate has the effect of reducing the parameter that
measures penetration even though the number of households competing for
available self-cut wood resources is steadily increasing. In the Portland
version penetration was defined as the ratio of households heating with wood
divided by the original number of households in 1970 (HWC/THI or households
with wood capacity/total households initial).
The model line numbers which incorporate these table values for the
Portland area are shown below:
1480 A ESCWP.K = TABLE (TESCWP, HWCK/THI, 0, 1, .2)
1490 T TESCWP = .4/.5/.8/1/1/1
(5) Fraction (of households) from Cost that Will Install Wood
Capacity (TFCIWC)
The TFCIWC variable represents the Fraction (of households) From Cost
(i.e. based on payback period) that will Install Wood Capacity. Based on the
A-16
-------
payback oeriod that a new wood installation represents a certain fraction of
households '-/ill install wood heating capacity. Since Portland area annual
heating costs are about half of New England costs, this author perceived that
sliqhtlv fewer Portland households would be motivated to install wood heating
even with the same payback period. Thus, a slightly lower installation
response rate was assumed for the Portland area. Figure 11 shows the table
function in graphical form for the two areas and Table 32 shows the
relationship in tabular format.
Year 2000 projected wood usage is exactly the same, regardless of the
Portland or New England values used, but the 1970 to 1980 Portland wood use
pattern fits known historical values better with the Portland curve, which
tends to confirm the hypothesis that the true relationship for Portland is
closer to the values in Column 3.
C. Fireplace Hood Usage Sector
As explained in the text, several factors are expected to motivate less
wood burning in fireplaces. 1) As fireplaces are converted to more efficient
stove installations, fireplace wood burning should drop. 2) If real wood
prices escalate rapidly, aesthetic burning of wood should also drop. Burning
in a simole fireplace at from negative to 20% efficiency does not have as
attractive a return on investment (in wood burned) as stoves.
In order to gain insight into how fireplace v/ood usage has been changing
in the late 1970's," it was decided to review several available surveys with
multiple year data. The best and largest available data base is a 5600
household survey, the New England Fuelwood Survey , conducted by the U.S.
Department of Agriculture, Economics, Statistics, and Cooperatives Service,
published in 1930. According to this survey, fireplace wood usage dropped by
18% between the 1976-77 and the 1978-79 heating seasons.
As additional corroboration of this trend, a comparison between the Talbot
Wong survey results for the Medford area for winter 1977-78 vs. an Oregon DEQ
survey for winter 1930-81 indicated an average annual wood usage per fireplace
of 2.15 cords for the winter 1977-78 period vs. 1.26 cords for the latter
period. Although the two survey methodologies and questions format are
slightly .different, this drop in average usage per fireplace confirmed the
expected drop in wood usage per fireplace.
Norman Marshall was consulted as to what might be the best factors to use
to project future fireplace usage. He suggested that the most simple and
straightforward method would be to have fireplace v/ood usage drop in proportion
to increases in average wood prices. A review of New England wood prices
during the two year period when an 18% drop in fireplace wood usage occurred
showed that prices for wood (in constant 1980 dollars) increased about 12%
during that same time period. The assumption that an elasticity relationship
of a 1.5% fireplace wood usage drop for each 1% price increase was considered,
but this (based on the 18% to 12% ratio) methodology would predict fireplace
wood usage dropping to zero if prices increased 67%. The New England wood
prices available and used in Marshall's model were actually New Hampshire wood
prices. The New England price increase could have been close to 18%. Thus, it
was decided to set future fireplace wood usage equal to base 1977 wood usage
times the ratio of 1977 wood price over future year wood price. In the example
for the Portland metropolitan area assuming a year 2000 wood price of $130/cord,
A-17
-------
Figure 11 Wood Installations vs. Payback Period
Fraction of
Households
Installing
Wood
Heating
Capacity
1
Marshal 1's Curve
10
Payback Period
(Years)
"A-18
-------
TABLE 32
Wood Installations vs. Payback Period
New England Fraction of Portland Area Fraction
Households Installing of Households Installing
Payback Period Wood Capacity Wood Capacity
0 .2 .2
2 .2 .18
4 .15- .12
6 .10 .07
8 .05 .03
10 0 0
12 0 0
14 0 0
16 0 0
18 0 0
20 0 0
A-19
-------
?009 Fireplace Wood Usaae = 1977 fireolace HOO^ usage (1977 wood price}
(2000 wood price)
= 219,000 (S85/S130) = 145,000 cords
Since no data was available on fireplace wood usage prior to 1977, the
simple assumption was made that 1970 through 1977 fireolace wood usage remained
constant. This methodology could he improved by mathematically linking a
certain fraction of new stove installations to a rate of fireplace
"decommissioning," but given the limited scope of the original task (which did
not include computer model projections) as well as the uncertainties about true
elasticities it was decided to incorporate this simple relationship.
The code below shows the code which was inserted into the Portland version
of Marshall's model. All additions to the code were inserted between lines
1700 and 1720 so that the line numbers of the balance of the Portland code in
the model still correspond exactly to line numbers used originally by Marshall
with this one exception. In order to change this code regarding input values,
the four occurrences of the value "209,000" (1977 fireplace wood usage) should
all be changed to whatever alternate initial fireplace usage is assumed, and
the 1977 wood price value in Line H06 (86, in this case) altered to the
aopropriate value.
1701 NOTE
1702 MOTE FIREPLACE USAGE MODULE
1703 NOTE
1704 NOTE
1705 A WPR.K=35/BWP.K
1703 A DWPR.K=DLINF1(WPR.K,FCSAT)
1710 N FWU2=209000
1712 A FUD.K=209000*(1-DWPR.K)
1714 A FWU.K=CLIP(FWU1.K,FWU2.K,1977,TIME.K)
1715 A FWU1.K=209000
1718 A FWU2.K=209000-FUD.K
1720 A TWU.K=FWU.K+WU.K
A-20
-------
2. SEATTLE MODEL DOCUMENTATION
In order to fully understand this section, the reader is advised to first
read the documentation for the Portland modeling work (in Appendix). This
section explains only modifications made to the Portland version of the model.
A. Local Condition Inputs
(1) Historical Wood Usage
Historical wood usage is discussed in Sections 2(D)(1) and
2(D)(2) of this Appendix, page A2-9 through A2-11.
12) Household Heat Requirement^
A direct estimate of average household heating requirements was not
available from either BPA, the Seattle Energy Office, or the Washington
Department of Natural Resources. Therefore, the heating requirements
determined for Portland were scaled based on heatinq degree averages. Since
King County ''Seattle) has heating degree days of 4447 (Washington Energy Use
Profile) compared to the Portland area average of 5036, the Seattle household
heating requirement was assumed to be 44.2 million BTU's/year (50 million x
447/5036). This is input in the model using the code below:
890 C HRHI = 44.2E6
(3) Household Growth Function
Estimates for households in the city of Seattle were available for 1970
and 1930. These were interpolated to derive a 1975 estimate. Population
orojections were available for the combination of King plus Snohomish County
(Seattle SMSA). These projection rates were used to project future households
after 1980 in Seattle. Table 33 shows the SMSA projections in Column 1, those
projections relative to 1.00 in Column 2, and resultant projections for the
number of households in Column 3.
A-21
-------
TABLE 33
Seattle Area Population and Household Projections
Year SMSA Population Growth Relative to 1.00 Households in Seattle
1970 NA NA 206,100
1975 NA NA 213,050
1930 1,506,495 1.000 220,000
1935 1,797,946 1.119 246,180
1990 1,959,967 1.226 259,720
1995 2,149,748 1.338 294,360
2000 2,360,356 1.469 323,180
(4) Initial Fraction of Households with Wood Heating
Data available from the 1970 Census of Housing indicated that .16% of
households relied on wood as their primary heat source in 1970. This was input
in Line 270 using the code below: _
270 c IFHWC = .0016
(5) Heating Requirements
Since Seattle area household heating requirements were determined to be
84.4% of Portland heating requirements, future heating requirements were set at
84.4?! of the future heating requirements estimates for Portland or 42.2% of the
requirements determined for New England households. This was input in Line
1050 using the code below.
1050 T THRH = 44.2/42.4/39.8/37.6/35.4/34.0/33.2
(6) Adjustment to Real Prices Based on the Consumer Price Index
All historical fuel prices were input into the model in 1930 constant
-dollars. Specific C.P.I, data" for the'Seattle area was obtained from the San
Francisco Office of the U.S. Bureau of Labor statistics (415-556-4678).
Seattle C.P.I, figures for 1970 to 1930 are shown in Table 34.
A-22
-------
TABLE 3*
Seattle Metropolitan Area Consumer Price Index 1970-1930
1970 - 112.5
1971 14.6
1972 119.0
1973 123.1
1974 135.8
1975 1.3
1976 161.7
1977 1.4
1978 184.1 '
1979 202.0
1930 235.0
(7) Historical Oil and Gas Prices
Historical oil prices were obtained from the OH Heat Institute of
Washington. Actual prices are shown in Table 35 in Column 1 with prices in
real 1980 dollars shown in Column 2. Historical gas prices were obtained from
Washington Natural Gas. Historical gas prices are shown in Column 3 with CPI
adjusted prices in Column 4.
TABLE 35
Historical Oil and Gas Prices in the Seattle Area
Actual Oil Price Oil Price, 1980$ Actual Gas Price Gas Price, 1980S
Year (S/gallon) (I/gallon) f$/lp6 BTU) fS/106
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1930
.206
.214
.227
.254
.344
.409
.463
.473
.495
.786
1.079
.432
.441
.450
.487
.598
.635
.676
.551
.635
.918
1.079
1.56
1.597
1.612
1.665
1.923
2.573
2.839
3.216
3.563
4.267
5.118
3.27
3.29
3.20
3.19
3.34
4.00
4.22
4.43
4.57
4.98
5.12
A-23
-------
(8) Historical Electricity Prices
Historical electricity prices were obtained from Seattle City Light in
Seattle. Actual prices are shown in Table 36, Column 1, with prices in real
1930 dollars in Column 2.
TABLE 35
Historical Electricity Prices in the Seattle Area
Year Actual Price (S/kwh) Real 1980 $ Price (S/kwh)
1970 .00862 .0181
1971 .00853 .0177
1972 .00858 .0170
1973 .00858 .0164
1974 .00922 .0150
1975 .00922 .0143
1976 .00922 .0135
1977 .01174 .0162
1973 .01174 .0151
1979 .01174 .0137
1930 .02069 .02059
(9) Historical Wood Prices
Historical wood prices for Seattle were obtained for 1970, 1975, and 1980
from the Seattle Post-Intelligence Classified Section. These values were
converted to 1980 real prices via the Consumer Price Index, and intermediate
year values derived via interpolation as shown in Table 37.
TABLE 37
Historical Seattle Area Wood Prices '
Year Actual Price (S/cord) Real Price (1980 S/cord)
1970 29 61
1971 62
1972 64
1973 66
1974 68
1975 45 70
1976 75
1977 80
1978 85
1979 90
1980 94 94
A-24
-------
The values in column 3 were input into the model using the code below:
1450 A TWPI = 61/62/64/56/58/70/75/30/85/90/94
(10) Historical and Projected Fuel Use Split
The 1970 Census of Housing provided information on the distribution of
oil, gas, and electricity usage 'as the primary heat source) in occupied
housing units. Those values are shown in Row 1 in Table 38. For the 1980 or
current fuel split for conventional fuels, the best available information
source is the Seattle City Light Survey (1979). 1930 fuel usage was assumed to
be split in the same proportion as from the S.C.L. data for fall 1979. ' This
information is shown in Row 3 below. The 1975 fuel split was derived by
interpolating between 1970 and 1930 and is shown in Row 2 below. For future
year fuel split projections, the only source obtained for the Pacific Northwest
was ODOE projections* for future state of Oregon fuel splits. Based on how
oil, gas, and electricity usage is projected to change in Oregon from 1980 to
2000, the Seattle area splits were projected to change in the same proportion
by 2000. The year 2000 derived split w« normalized to sum to 1.00 and values
for years 1985, 1990, and 1995 derived by interpolation. The 1985-2000 values
for each fuel share are shown in Rows 4 through 7.
TABLE 38
Historical and Projected Fuel Use Split in the Seattle Area
(Relative to 1.00)
Year - Oil Use Gas Use Electricity Use
1970 .530 .249 .221
1975 .460 .240 .300
1980 .389 .230 .481
1985 .351 .223 .501
1990 , .314 .216 .521
1995 .276 .208 .541
2000 .238 .201 .561
*ODOE Oregon's Energy Future, op. cit., p. 100
Only the values for oil and gas are input into the model with electricity
shares derived by difference. Lines 1080 and 1100 input the oil and gas splits
in the model, respectively, as shown in the code below:
1030 T TFCFO = .537.46/.3897.351/.314/.2767.238
1100 T TFCFG = .2497.24/.237.2237.2167.2087.201
* A-25
-------
ons
i n
(11) Future Oil and Gas Prices
Future oil and gas prices were based on real price growth rate projecti
for the Pacific Northwest by BPA.* These projected real prices are shown i r
Table 39. Values are shown in 2.5 year increments to be consistent with model
input requirements.
TABLE 39
Projected Real Prices for Oil and Gas in the Seattle Area
Real Oil Price Real Gas Price
Year (1930 S/gallon) fl980 $/lQ6_BTUl
1980 1.079 5.12
1982.5 1.200 5.93
1985 1.343 6.73
1987.5 1.430 ~ 7.36
1990 1.538 8.01
1992.5 1.665 8.74
1995 1.812 9.48
1997.5 1.880 10.24
2000 1.964 11.03
The data in Columns 2 and 3 above are input into the model using the code
below:
1190 T TOP2 = 1.079/1.20/1.343/1.43/1.538/1.655/1.812/1.88/1.964
1280 T TGPZ = 5.12/5.93/6.73/7.36/8.01/8.74/9.48/10.24/11.03
(12) Future Electricity Prices
The real electricity price growth rate projections developed by BPA. for
public utilities in the Pacific Northwest** were used to project future
electricity rates for Seattle. The escalation factors and resultant
projections are shown in Table 40:
*BPA, Economic, Demographic Projections of the Pacific Northwest, op. cit.,
p. 147.
**BPA Pacific Northwest Electric Energy Model, op. cit.
A-26
-------
TABLE 40
BPA Electricity Price Growth Rates and
Resultant Seattle Area Projections
Growth Rate Projections for Seattle Area Projected
Year Public Utilities (1980=1.000) Electricity Prices (S/kwh)
1980 1-000 .0208
1982 5 1.078 .0223
1935' 1.155 .0239
1937 5 1.291 .0267
1990 1.427 .0295
1992 5 1.473 .0305
19-95 1.519 .0314
1997 5 1.5825 .0327
2000 1.646 .0339
The values in Column 3 are input into the model using the code below:
1370 T TEPZ = .0207/.0223/.0239/.0267/.0295/.0305/.0314/.0327/.0339
(13) Future Wood Prices
As for the other cities in this projection study, a 2% real wood price
escalation factor was assumed and results input into the model using the code
below for years 1980 through 2000 in 2.5 years increments.
1470 T TWPZ = 94/99/104/109/114/120/126/132/139
B. Explanation of Table Functions Adjusted
Only two of the "relationship table functions" were changed for the
Seattle runs. Since heat requirements are lower, the same stove installation
represents installation of a higher fraction of potential wood heating capacity
(closer to 1.00). Thus, for a given fraction of capacity installed (i.e., 60%
of household heating requirements) the inconvenience of wood heating should be
slightly lower for that fraction of capacity and the wood stove installation
cost slightly lower (as compared to the wood use and stove size needed to
provide 60% of heating requirements in Portland or New Engla.nd). The two
functions changed were TCIWC and TNWIC.
(1) Cost of the Inconvenience of Wood from Capacity (TCIWC)
This function expresses how the cost of the inconvenience of wood due to
capacity increases as the capacity fraction installed increases. As explained
for TCIWC in the Portland section of the Appendix, the New England table values
A-27
-------
were simply ratioed downward in proportion to the ratio between Seattle
household heating requirements (44.2 x 105 BTU's/year) and New England heating
requirements (100 x 105BTu"s/year) in 1970.
The resultant code is shown in line- 1000 below:
1000 T TCIWC = 83/133/220/354/440/440
(2) Normal Wood Installation Cost (TNWIC)
Since heat requirements are less, the size of stove (and therefore cost)
that must be purchased to provide 100% capacity is therefore less. The cost of
100% capacity was adjusted by the ratio of Seattle to New England Heating
Requirements ($1500 x 44.2/100), and the cheapest stove available was still
assumed to be $200. Intermediate values were interpolated based on the
proportions in the original New England function. The resultant code is shown
below:
480 T TNWIC = 200/235/270/340/470/660
C. Fireplace Wood Usage Sector
The exact same code used in the Portland version of the model was used for
Seattle, with the exception that 1977 fireplace wood usage was set at 100,000
cords/year. The value "100,000" replaced the Portland fireplace wood usage
value of 209,000 in Model Lines 1710, 1712, 1716, and 1718.
D. Seattle Model Calibration
Calibration of the model for the Seattle area is more difficult because
there is no complete survey of wood usage available. The available survey
information consists of the following:
o A Seattle City Light (private utility) survey which provides
information on the number of fireplace and stove installations in all
households in Seattle, but did not question, respondents about the
amount of wood they burn. The S.C.L. service territory is primarily
composed of the city of Seattle.
o A methodologically complete survey conducted by Del Green which provides
good wood usage information for both stoves and fireplaces, but is questio
ably representative of the city of Seattle because it only surveyed 800
contiguous households within one neighborhood outside of the city limits.
o A Puget Sound Power and Light Survey that provides information on the
number of fireplace and stove installations in households, but
did not question respondents about the amount of wood they burn.
Further, the P.S.P.L. service territory covers all of King County
excluding the Seattle City Light territory (and therefore excludes
Seattle).
The primary purpose of this task was to project wood usage for Seattle.
The projection was made by developing 1980 estimates of wood usage for the City
A-28
-------
in the Seattle area is
City Light) provides
city of Seattle or the
does not quantify wood
of Seattle based on information from the first two surveys cited above. Future
projections would be based on those values. The following section discusses in
more detail how the 1980 wood usage estimates were derived from these two
surveys.
Available survey information on wood usage
incomplete. One private utility survey (Seattle
information for an area slightly larger than the
fraction of households with stoves and fireplaces but
usage per household. The survey conducted by Del Green provides the needed
wood usage data but only covers 800 contiguous households in one neighborhood
and is therefore questionably representative of even the city of Seattle.
However, no better information is available. Wood usage estimates were derived
using these two surveys and averaged in an attempt to minimize error. The
categories of wood usage for heating purposes and "aesthetic" purooses are
considered separately.
(1) Wood Usage for Heating Purposes
(a) Seattle City Light Surrey
The S.C.L. Fall 1979 survey determined the number of households that 1)
uses wood stoves/furnaces as a primary heat source, 2) use wood stoves/furnaces
as a secondary heat source, and 3) use fireplaces with heatilators. Since no
data is available on wood usage in these categories of households, the usage
rates from the Del Green survey were applied yielding an estimate of wood
burning for heating purposesas shown below.
Burninq Unit
Stoves-Primary
Heat Source
Stoves-
Secondary
Heat Source
Fireplaces
Heatilators
Seattle
(City)
Households
220,000
220,000
220,000
Seattle City
Light Survey
Based Fraction
of Households
with these units
.009
.033
.041
Del Green
Survey-Based
Wood Usage
Estimates
(cords/HH/hr)
3.19
1.58
2.94
Resultants
Wood Burninq
(cords/year)
5316
11471
26519
44306
cords/yr
(b) Del Green Associates, Inc. Survey
The Del Green survey of a Bellevue neighborhood provided information on
the fraction of households with wood burning units and on the amount of wood
A-29
-------
usage ner unit. This information is shown following:
Del Green
1930 Survey-Based Del Green
Seattle Fraction of Survey-Based Resultant
Burning (City) Households with Wood Usage Wood Burning
Unit Households Units (cords/HH/yr) (cords/year ^
Furnaces and 220,000 .040 3.19 28072
Stoves-
Primary Heat
Source
Furnaces and 220,000 .032 1.58 28503
Stoves-
Secondary Heat
Source
Fireplace 220,000 .019 - 2.94 12289
with "insert"-
Primary
Usaqe
68S64
The average amount of wood burning in 1980 for "heating" purposes from
these two surveys is 56,600 cords per year. Clearlv, there is a lot of
uncertainty remaining until a survey is conducted over a larger geographical
area and requests information on both the types of units and amount of wood
burned in those units.
Seattle City Light is planning another wood burning related survey in the
summer of 1981. The utilities need this information if they are to develop
precise estimates of how much their electric space heating load would be
displaced by households relying on wood space heating. Air pollution agencies
would be well advised to-cooperate with the utility so that wood usage
questions can be added to the survey. The agencies should desire this data to
determine the amount of wood burned in stoves vs. fireplaces (since pollutant
emission rates are different)
(2) Fireplace-Aesthetic Wood Usage
(a) Seattle City Light Survey
The S.C.L. Survey found that 26.7% of households in their service
territory have fireplaces without any heatilators, etc. Based on the Del Green
survey finding that such units burned about .7 cords/year, an estimate of
fireplace wood usage of 41,100 cords is derived. The S.C.L. survey found that
4.1% of households they served had fireplaces with heatolators. Assuming the
rate of usage from the Del Green Associates, Inc. survey (1.76 cords/HH/year)
for these units affords an estimate of 15,900 cords/year. Thus, total fire-
pi aca wood usage is estimated to be 57,000 cords/year.
A-30
-------
(b) Del Green Associates, Inc. Survey
The remaining three categories of wood hurninq from the Del Green Survey
are shown below, together with wood usage estimates in those categories.
Burni ng
Unit
Fireplaces -
No Inserts,
Primary Usage
Fireplaces-
with Inserts
Secondary
Usage
Fireplaces-
No Inserts
Secondary
Usage
Fireolaces
with Inserts-
Primary
Usage
1930
Seattle
(City)
Households
220,000
220,000
220,000
220,000
Del Green
Survey-Based
Fraction of
Households with
Uni_ts_
.013
.1*9
.651
.019
Del Green
Survey-Based
Wood Usage
(cords/HH/yrl
1.38
1.51
0.70
2.94
Resultant
Wood Burning
(cords/year^
3947
52776
100254
12289
169266
The average of these two survey-derived estimates is a fireplace wood usage
estimate of 113,000 cords/year for the city of Seattle for 1980.
(3) Comparison of Model Predicted 1970-80 Wood Usage with Survey
Results
Local input factors such as Seattle area household heat requirements, fuel
prices, etc. were input into the model. Additionally, several of the functions
which are influenced by the annual heating requirements of an area were
modified, as explained in Section 2b of this Appendix.
Based on the review of survey information, 1980 wood usage by the 220,000
households is estimated to be about 42,000 cords per year in stoves and
furnaces and about 113,000 per year in fireplaces. The rate of wood usage for
primary heating purposes appears to be lower in the Seattle area than in the
Portland area. This is hypothesized to be largely due to Seattle's lower
electricity rates Cslightly over 20 mills/kwh vs. 37 mills/kwh in Portland in
1980). Figure 12 shows the best estimate of wood usage trends in Seattle City
between 1970 and 1980?. There is an acceleration in wood .usage between 1976
and 1980 as would be expected.
Based on input factors which are representative of Seattle area
conditions, 1980 wood usage in stoves or for primary heating purposes was
A-31
-------
FIGURE 12
Model Estimated Stove Wood Usage in Seattle City
1970 - 1982 (Cords/year)
90,000
60,000
30,000
1970
1975
1980
A-32
-------
v.V
predicted to be about 39,000 cords -per year. The 1979 and 1981 v/ood usage in
stoves is estimated by the model to be 57,000 and 57,000 cords/year,
respectively. This prediction by the model was considered to be in agreement
with the survey derived estimate of wood usage of 42,000 cords/year in 1980.
Regarding total wood usage, the annual growth rate projected by the model
between 1979 and 1982 is 6* per year, which is consistent with the short term
trend findings in Section 2 of this report. Stove, fireplace and total wood
usage is shown in Figure '13.'
Seattle City Light will be conducting another wood heating survey in the
summer of 1981. If questions can be incorporated into the survey on wood usage
rather than just on the fraction of households with stoves and fireplaces, the
additional information should be of value to the utility (to project how much
heat is being displaced by wood heating) as well as to air pollution agencies
(interested in wood usage in fireplaces and stoves separately, because of their
different emission factors).
(4) Base Case 1970-2000 Projected Wood Usage in Seattle City
The projections through the year 2000 indicate the Seattle total wood
usage will increase about 2Q% in the 1980 to 1985 period, remain about constant
through 1993 and decrease about 102 in the period following through 2000. The
model projects a peaking of wood combustion in the latter 1980's with a
decrease projected for the 1990-2000 decade, unlike the acceleration in wood
usage projected for the Portland area for that period.
This is believed to he attributable to several factors:
o Seattle heating requirements are about 20% less than Portland
reducing the potential savings from wood heating
o Similar electricity price escalation factors (from SPA) are applied
for both cities, but the baseline 1980 rates are about 435: lower for
Seattle. Therefore Seattle rates are about 43* lower in all future
years. Households heating with electricity have less savings
potential from heating with wood and therefore are less likely to
convert to wood.
o Lesser population growth is projected for the Seattle area.
The model projections for wood usage in the Seattle area are shown in
Figure 14.
Conclusions based on these projections are discussed in the Findings and
Conclusions section of the text (Section III).
A-33
-------
CO
*
-H TI
O ->
rt -S O
O) n> <
(/) O
D< fD
IQ
fD
Cords Wood/Year
CO
O
I*
O
O
O
CTi
O
O
O
O
VD
CD
u
O
O
O
ro
o
v
o
o
o
O
O
O
O>
o
CD
o
o
I
CO
n
o>
-s
UD
oo
o
Jl-
co
(D
01
rt
-s
ro
01
o -
i -3
UD
rjo
ro
o
o
-5
cx
in
O
Q-
fD
Ot
rt
fD
CX
O
O
CX
in
O>
UD
n>
CO
fD
01
-5
3=-
T3
Ol
13
n
ro
00
-------
Cords Wood/Year
a
-J -O l/>
O ' rt
rt -t O
O fD <
'-O fD
_^
c= o>
in o
& n>
ca
ft>
CO
en
ro
Dl
-j
--J '
o
1
ID
oo '
o
-
,_,
10
UD
O
tsj
O
o
0
i f rx
tn o en c
0 0 O C
o o o c
0 O 0 C
o o o c
.
9
1
1
1
1
1
1
1
O
O
a
u
m
m
i
i
i
i
i
i
i
!
a
a
a
a
a
a
D
a
a
a
D
D
a
a
a
a
o
a
a
a
a
a
00
ro
ai
o
I.
rt
CO
ai
(/i
fD
o
ai
^
1
UD
1
O
|
no
o
o
o
>»
o
Q-
ro
'
' ~U
^ ~5
O
C-j.
CO
O
c-t-
fD
a.
en
c:
o
o
-s
Q.
10
fD
CU
-5
O
O
O-
in
at
to
fD
-------
3. SPOKANE MODEL DOCUMENTATION
In order to fully understand this section, the reader is advised to first
read the documentation for the Portland modeling work (in Appendix). To the
extent that some model impacts for Seattle were also used for Spokane, the
reader may also need to review certain sections of the Seattle documentation,
but all these instances are specified in this section. This section explains
only modifications that are different from the cod« used for either Portland or
Seattle.
A. Local Condition Inputs
(1) Spokane Area Historical Wood Usage
Available survey information on wood usage in the Spokane area is
incomplete. A Del Green Associates, Inc. survey is available covering
one neighborhood in the Spokane area. Respite questions about the degree
to which it is representative of conditions in the entire City of Spokane,
it was judged to be the best available source of information. The
Washington Water Power Survey covered a much larger territory and did not
provide information on wood usage per wood burning unit. The table below
shows the derivation of the estimates that 1980 wood usage in Spokane was
about 26,000 cords in stoves and about 96,000 cords in fireplaces.
Burning Unit
Type
Stoves-Primary
Stoves-Secondary
1930 Spokane
(City)
Households
70,900
70,900
70,900
70,900
70,900
Fireplace/Secondary 70,900
Fireplace Insert
Primary
Fireolace Insert
Secondary
Fireplace/Primary
Fraction of
Households
With Unit
.025
.032
.028
.260
.059
.503
Average 1980 Resultant
Use Wood Burning
(cords/HH/yr) cords/yr)
4.98
2.98
4.3
2.33
2.37
.94
8,827
17,325
11,594
33,323
96,404
(2) Household Heat Requirements
Since a direct estimate of household heating requirements was not
available for Spokane, the Spokane requirement was estimated by ratioing
Spokane heating' degree davs to Portland heating degree days. The 1970 initial
household heating requirement ws estimated to 64.7 million BTU's per year
(6518/5035 x 50 x 10° = 64.7 x 106). This was input in the model using the
code below:
890 C HRHI = 50E6
A-36
-------
(3) Household Growth Function
Estimates of the number of households in the city of Spokane in 1970 and
1930 were available from U.S. Census Data. A 1975 value was derived by
interoo1 ation . For future year projections, the Sookane SMSA growth
orojections were applied to estimate growth in the number of households after
1980. The table following shows the household projections in Column 4 and the
basis for projecting years 1985, 1990, 1995, and 2000 in Columns 2 and 3.
Year
1970
1975
1930
1935
1990
1995
2000
SMSA
Peculation
Projections
NA
MA
337502
371000
399733
432529
470530
SMSA Growth
Relative to
1.00
1.000
1.099
1.184
1.282
1.394
Spokane City
Number of
Households
5044?
65579
70916
77937
33955
90914
98357
The values in Column 4 of the above table were converted into annual
growth rates for each of the five-year periods between 1970 and 2000 and input
the model via the code below:
in
300 T THGF= .015/.016/.019/.015/.016/.017/.017
Initial Fraction of Households with Wood Heating
indicated that .39% of
in 1970. This was input
Data available from the 1970 Census of Housing
households Belied on wood as their primary heat source
in the model using the code below:
270 C IFHWC = .0039
(5) Heating Requirements
Since Spokane area heating requirements were determined to be 129.4% of
Portland requirements, the projections for how heat requirements will drop
through the year 2000 were set at 129. 4£ of the Portland values, and the
following code was input in the model.
1050 T THFH = 64.7/62.1/58.2/55.0/51.8/49.8/48.5
A-37
-------
(6) Adjustment to Real Prices Based on the Consumer Price Index
All historical fuel prices were input into the model in 1930 constant
dollars. Since year by year CPI fiqures were not available for the Spokane
area, Seattle CPI figures were utilized. These were documented in the Seattle
section of this Apoendix.
(7) Historical Oil Prices
The Oil Heat Institute of Washington stated that historical oil prices for
Spokane and Seattle were essentially the same. Thus, the same code was used
for Spokane as for Seattle for historical oil prices.
(8) Historical Gas Prices
Historical gas prices in the Sookane area'were obtained from Washington
Water Power and are shown in Column 2 of the following table. Gas prices in
1980 dollars were derived by the CPI, jnd the resulting historical real p-ices
for gas are shown in Column 3 below.
Year Actual Gas Price f$lQ5 BTlM Real 1980 S Gas Price CS/1Q6 BTU)
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1.07*
1.155*
1.165*
1.222*
1.737*
2.079
2.650
2.754
3.033
3.947
4.478
2.24
2.40
2.31
2.34
3.02
3.23
3.87
3.79
3.39
4.61
4.48
*Rates changed after first 65 therms used in a month
The values from Column 3 above were input in the model using the code
below:
1260 T TCPI = 2.24/2.4/2.31/2.34/3.02/3.23/3.87/3.79/3.89/4.61/4.48
A-38
-------
(9) Historical Electricity Prices
Historical electricity (actual) prices were obtained f<"om Washington water
power and are shewn in Column 2 below. Prices in real 1930 dollars were
derived using the CPI and are shown in Column 3.
Sookane Area Historical Electricity Prices
Year
1970
1971
1972
1973
1974
1975
1976
1977
1973
1979
1930
Actual Electricity
Price f$/!
-------
:n
Historical and Projected Fuel Spl>t
The 1970 census of housing provided information on the distribution of
oil, qas, and electricity usage fas the primary heat source) in occupied
housina units. For 1980, there is no survey information that has geographical
coverage comparable to Spokane. The Del Green Survey results cover an area of
300 contiguous households and a utility survey by Washington Water Power
provides information for an area larger than the SHSA. The fuel splits from
these two information sources are significantly different, and there is fuel
solit information on an SMSA basis for both 1970 and 1977. The table below
shows the different fuel splits which are available for years 1970 and 1980.
Spokane Area Fuel Splits Crelative to 1.00)
Source
Del Green
Survey
Washington
Water Pov/er
Survey
Census
Census
Census
Area Coverage Oi1
Neighborhood in .034
Spokane
1980
Larqer than
Spokane SMSA
1970 Sookane City
1970 Spokane SMSA
1977 Spokane SMSA
.24
.43
.40
.287
Gas
.832
.39
.47
.44
.421
Electricity
.034
.23
.10
.15
.292
Two methodologies were emoloved to develop 1980 fuel splits for Spokane.
First the Del Green and Washington Water Power splits for 1980 were averaaed.
Second, the assumption was made that the SMSA trends between 1970 and 1977 also
occurred in Spokane. This trend was extrapolated through 1980. The first two
rows in the table following show the splits derived by these two methods. Note
the electricity heating share is about the same. This minimizes concern about
developing the correct splits since oil and gas prices tended to increase over
the 1970-1980 period at similar rate. The model calibration of heating costs
are not that different when different oil vs. gas splits are input. For the
assumed 1980 fuel split in Spokane, the first two rows in the table were
averaged and assumed to represent the 1980 Spokane fuel split. 1975 values
were derived by interpolation.
Source Year
Average of Del 1980
Green * WWP
Spokane City 1980
projected to
1980 based on
70-77 SMSA
Trends
Average of 1980
Above Two
Estimates
Oil
.16
.28
Gas
.61
.48
Electricity
.23
.24
.22
.545
.235
A-40
-------
For future year fuel sol it projections, the only source obtained for the
Pacific Northwest was ODOE projections for the future fuel splits for the state
of Oregon. Based on how oil, gas, and electricity usage is projected to change
in Oregon from 1980 to 2000, the Spokane area splits were projected to chanqe
in the same proportion by 2000. The year 2000 derived split was normalized to
sum to 1.00 and values for years 1985, 1990, and 1995 derived by interpolation.
The 1935-2000 values are shown in rows 4 through 7 in the table below.
Historical and Projective Spokane Citv Fuel Splits
(Relative to 1.00)
Year Oil Gas Electricity
1970 .43 47 .10
1975 .325 .508 .157
1930 .22 .545 .235
1935 .203 .543 .254
19<>0 .185 .541 .272
1995 .169 .539 .291
2000 .152 .537 .309
Onlv the values for oil and gas are input into the model, with electricity
shares derived by difference. Lines 1030 and 1100 impact the oil and gas
snlits in the model, respectively, as shown in the code below:
1080 T TFCFO = .43/.325/.22/.203/.185/.169/.152
1100 T TFCFG = .47/.508/.545/.543/.541/.5.19/.537
(12) Future Oil and Gas Prices
Future oil and gas prices were based on real price growth projections for
the Pacific Northwest by BPA. These projected real prices are shown in the
table below. Values are shown in 2.5 year increments to be consistent with
model inout requirements.
Projected Spokane Area Oil and Gas Prices
Year Real Oil Price (1980 S/aallon) Real Gas Price (1980 $/106_BJUl
1930 1.079 4.48
1982.5 1.200 5.18
1985 1.343 5.89
1987.5 1.430 6.45
1990 1.538 7.01
1992.5 1.665 7.65
1995 1.812 8.30
1997.5 1.880 8.93
2000 1.964 9.66
The data in Columns 2 and 3 above is input into the model using the code
below:
1190 T TOPZ = 1.079/1.2/1.343/1.43/1.538/1.655/1.812/1.830/1.964
1230 T TGPZ = 4.48/5.18/5.39/6.45/7.01/7.55/8.30/8.98/9.65
*BPA, Economic, Demograohic Projection of the Pacific Northwest, op. cit., p.
147.
A-41
-------
BPA Electricity Price Growth Rates and Resultant Spokane Projections
Growth Rate Projections for Spokane Area Projected Real
Year Public Utilities f1980=1.0001 Electricity Prices (S/'
-------
(2) Normal Wood Installation Cost (TNWIC)
Since heat requirements are less, the size of stove, and therefore cost,
that must he purchased to provide 100% caoacity is therefore less. The cost of
100*o caoacity was adjusted by the ratio of Sookane to New England Heating
Requirements (51500 x 64.7/100), and the cheapest stove available was still
assumed to be S200. Intermediate values were interpolated based on the
proportions in the original New England function. The resultant code is shown
in Line 480 below:
480 T TNWIC = 200/259/317/434/570/970
C. Fireplace Wood Usage Sector
The exact same code as used in the Portland and Seattle versions of the
model was used for Sookane, with the exception that 1977 fireplace wood usage
was set at 100,000 cords/vear. The value "100,000" replaced the Portland
fireoface~wood usage value of 209,000 in-Model Lines 1710, 1712, 1715, and 1718.
D. Spokane Model Calibration
(1) Wood Use Survey Information
Calibration of the model for Spokane is more difficult than for Portland
because no complete survey of wood usage is available. Two surveys with
potential were available and these included:
o A Del Green Associates, Inc. survey17 of a neighborhood in Spokane
with information oroduced on the amount of wood usage in each of six
different appliance categories. T^e chief problem with this survey
is its questionable representativeness of Spokane because the sample
is from geographically small area.
o A Washington Water Power wood usage survey!9 for a large service
territory including Spokane. The two chief problems with this survey
are: 1) wood usage information is only available in terms of the
"fraction of households relying on wood as the primary heat source";
and 2) the "Spokane" division of WWP's service territory is three
times as large as the population in Spokane.
Given these shortcomings in applicability for the utility survey, 1980
wood usage estimates were based on the Del Green Associates, Inc. survey
results." Projections were for the city of Spokane as originally required since
there was no good AQMA scale data base. The resultant estimate is that
stove/furnace wood usage was 26,000 cords/year and fireplace usage was 96,000
cords/year in 1980. Appendix A discusses the derivation of these values for
Spokane based on survey information.
A-43
-------
(2) Comparison of Model Predicted 1970-1980 Wood Usage with
Survey Results
Figure 15 shows the model-generated estimate of wood usage in Spokane
between 1970 and 1980. There is an expected acceleration in wood usage between
1976 and 1980. Based on input factors, which are representative of Spokane
conditions, 1930 wood usage was predicted by the model to be about 28,000
cords/year. Given the derived estimates of 1980 stove wood usage of 26,000
cords/year and the model prediction of wood use acceleration between 1976 and
1930, the Sookane version of the model was considered to be calibrated.
Regarding total wood usage, the annual growth rate projected by the model
between 1979 and 1981 is 6.4% per year, which is quite consistent with the
short term trend findings in Section II of this report. Stove, fireplace and
total wood usage between 1970 and 1932 is shown in Figure 15. Fireplace usage
is' set at 95,000 cords/vear initially.
(3) Base Case 1970-2000 Projected Wood Usage in Spokane
Projections through the year 2000 predict that total wood usage will level
in the early 1980's and begin to decrease in the latter part of the 1230's
(about 1936) at about a -.3% annual rate through the year 2000. This is
believed to be attributable to several factors:
o Sookane electricity rates are still low relative to most other
locations. For example, 1980 and 1981 rates at 2000 kwh/month were
1.45 and 1.98C/kwh. When BPA escalation factors are apolied, future
electricity prices are lower relative to other areas, and the savings
potential from wood heating is less.
o Regional population growth between 1980 and 2000 is projected to be
39% compared to 56% for the Portland metropolitan area. If Spokane
were to grow by 56% between 1980 and 2000, its total wood use would
be projected to remain essentially constant between 1980 and 2000.
The model projections for wood usage in Spokane are shown in Figure 16.
A-44
-------
^?s
«-* -j o
Co 3 <
O
c: n>
in
fu
to
Cords Wood/Year
OJ
o
i*
o
o
o
CTl
O
M
O
O
O
10
O
V*
o
o
o
o
o
o
o
I
-Pi
cn
vo
--J
cn
fD
tu
-s
00
O
" *
» *
o
a.
n>
CO
a
o
7T
01
3
rt>
o
. _..
(- 3
(JD
OO
n
o
o
o
QJ
O>
cr
ro
cu
TD
a
01
3
O
TO
to
-------
Cords Wood/Year
O c r+
rt- -$ O
CU 3 <
' & fD
n>
10
fD
CO
o
~C3
O
o
en
o
o
O
o
10
o
CD
o
o
§
O
en
CD
CD
O
O
IO
-^r
o
co
o
-P.
en
m
CU
10
10
CD
ro
o
-o
o
7T
CU
m
o
in
rD
o
CU
o
ro
O
O
0
^ ~
-a
-> -$
3 O
C_i.
fD
O
ri-
fD
CL
CT>
cz
m
i »
01
o
o
o
o
Q-
fD l«
CU O)
1 IQ
' fD
-------
APPENDIX B
Major Factors Not Included in Marshall's Model
-------
Marshall's model is driven to a large extent by the relative costs of
the different energy forms. A number of other factors not included in the model
can substantially affect wood burning levels. These include limitations on
wood supply, the effect of conservation, and possible government actions affecting
wood burning levels.
Marshall recognized that one of the weaknesses of his model is a lack of
input for factors influencing wood supplies. Predicting the demand placed on
wood residues from competing users, the resultant prices each could pay, and
the interaction of the competing useswouid in itself require substantial research
which was beyond the scope of Marshall's study on New England and the scope of
this study. Marshall (for New England) and the authors of this document chose
a 2%/year real increase in fuel wood prices, which approximately corresponds
with the expected increase of conventional fuels. Information follows on factors
influencing wood supplies for the readers' further understanding of this topic.
Availability of Wood Resources
In order to assess the feasibility of a 50 megawatt wood-fired power plant
in th.e Pacific Northwest, the Bonneville Power Authority conducted a literature
review of timber supply data to develop estimates of logging residues available
in 14 geographical areas within Oregon and Washington. This work represents
the best available estimate of logging residues which could be available for use
in residential stoves and fireplaces. Figures 17 and 18 show the annual avail-
ability of those logging residues as estimated in 1980 for the states of Oregon
and Washington.
Over the next 20 years, the utilization of much of the remaining old-
growth timber stands and the increased reliance on young-growth stands is
B-l
-------
CD
ro
FIGURE 17
Oregon Unused Wood Residue Quantities
(Annual Quantities in Dry Tons)
47
76,800 - Mil
409,200 - Fores
N. Willamette
28,600 - Mill
158,700 - F.
Mid-Willamette
0 - Mill
173,000 - Forest
Bend Prineville
45,900 - Mill
519,100 - Forest
12,100 - Mill
413,300 - Forest
37,200 - Mill
370,600 - Forest
93,700
(Mi 11)
239,800
(Forest)
11,900 - Mill
676,200 - Forest
79,800 - Mill
480,000 - Forest
Blue Mountains
125,700 - Mill
630,000 - Forest
-------
FIGURE 18
Washington Unused Wood Residue Quantities
(Annual Quantities in Dry Tons)
47
CO
I
CO
Puget Sound
Inland Empire
73,400 - Mill
562,500. - Forest
103,100 - Mill
815,400 - Forest
Olympia Peninsula
193,000 - Mill
1,168,700 - Fores
Central Washington
24,800 - Mill
1,193,800 - Forest
Lower Columbia
85,200 - Mill
551,100 - Forest
-------
likely to reduce the available mount of logging residues. Harvesting old-
growth stands results in considerably more residue than young-growth stands
due to the greater percentage of wood in branches and tops and because the
old-growth trees did not grow under managed conditions. According to the
42
Washington State Department of Natural Resources' Wood Waste for Energy Study ,
second growth timber stands generate 70 to 90% less forest residues during
harvesting than old growth stands. This could be partially offset by wood
from intensified management of second growth stands (i.e., wood from thinnings).
43
A 1980 Oregon Department of Forestry report projects that while 46% of Oregon's
timber harvest volume was derived from trees of 21 inches diameter or greater
in 1980, the percentage of the harvest volume from this class of trees in the
year 2000 is projected to be only 26%.
A potential counterbalancing trend may be increased removal of firewood
from forest lands held by individuals. As an example, in 1979, a pilot fuel
wood program was tested in New England under the Agricultural Conservation
44
Program of the Agricultural Stabilization and Conservation Service
Assistance was used to mark trees for cutting and to provide access roads
into the stand. Landowners could then sell the marked trees for firewood. The
first year program was so successful it was expanded to five states in 1980 and
further expansion is planned. The 1980 program cost about $2 million and
produced some 250,000 tons of firewood, which is about $8 per cord. This sort of
program could result in more firewood becoming available in the Pacific Northwest.
Demand for Wood for Other Uses
Demand for forest products for uses other than firewood is projected to
increase rapidly through the year 2000. The most comprehensive set of projections
3-4
-------
for these other uses are presented in the US Forest Service document entitled
45
An Analysis of the Timber Situation in the United States 1952-2030 . A
summary of the projected demand for various end uses of wood is shown in Table 41
TABLE 41
Projected U.S Growth in Demand
for
Various Timber Products, 1976-2000
1976
2000
Increase
Lumber
(109 bd.ft.)
42.7
59.9
40%
Plywood
(3/8" basis,
109 sq. ft.)
20.6
30.0
46%
Fiber board
(3/8" basis,
109 "sq.ft.)
13.5
25,3
87%
Other Fibrous
Materials
106 tons
60.2
121.5
102%
Roundwnod
rio9 cu.ft.)
13.3
22.7
71%
These projections indicate that unless total wood supplies are greatly
expanded, there will be increasing competition among the various end uses of
wood that compete with the use of wood for firewood. The largest increases are
projected for those wood uses (fiberboard and other fibrous materials) that
compete most strongly with firewood because they can utilize wood residues and
mill wastes.
41
At the Third Annual National Biomass Systems Conference in 1979 , Charles
Hewett noted that "The depletion of old growth timber (stands in the Pacific
Northwest) will increase the demand for the production of particle boards, fiber-
boards, and similar products as substitutes for lumber and plywood. Indeed, the
production and consumption of such products is already rising rapidly (USDA 1973,
USDA 1979). The raw materials for these products are very similar to the wood
B-5
-------
fiber used for energy production. With the rapid development of new processes
for the production of these materials, the forest products industry will
obviously compete increasingly with the wood-energy market during the next
41
thirty years." Unfortunately, Marshall's model does not take into account
the effect of limited availability of fuel wood because of competition by other
users.
Conservation Measures That Reduce Home Heating Energy Demands
A variety of actions can be undertaken to reduce household heating demand.
Since the driving force behind the conversion of household heating systems to
wood heat is the desire to reduce space heating costs, these various conservation
actions can reduce the need and desire of households to switch to wood to save
money. These actions can be grouped into four major categories:
Household Weatherization
Fireplace Modification
Localized Room Heating
Humidifi cation
Weatherization
Household weat.herization, which includes actions such a:; increased
insulation of roofs, walls and flooring, double and triple window glazing,
weatherstripping and caul king, can reduce household air leaks and thereby the
amount of energy required to maintain comfortable temperatures. The Portland
Energy Office cites reductions in space heating of 30 to 50% as achievable for
poorly insulated homes . In the Oregon Department of Energy's publication
Weatherization, One Step At A Time , Governor Victor Atiyeh states that "60%
of Oregon's homes are poorly insulated. If all our homes were properly
B-6
-------
protected from heat loss, we could cut our household heating requirements by
as much as 30%".
For new homes with the most sophisticated weatherization technology, the
reduction in space heating requirements that is achievable is astounding. A
January 1981 article in Canadian Energy News cites weatherization techniques
that can reduce annual heating costs in 9000 degree-day climates to under $100
57
per year .
Although weatherization can significantly reduce heating costs, a sig-
nificant deterrent to implementation of weatherization actions are the initital
costs, which often exceed $1000. Even though this weatherization investment
can produce permanent savings after a reasonably short payback period, many
homeowners tend to compare the initial weatherization costs to the annual costs
of continuing to heat without weatherization rather than comparing these two
costs on an appropriately discounted basis. Several utilizites and municipalities
throughout the Pacific Northwest have established grant and low interest loan
programs to help overcome individuals' aversion to high initital first costs.
Some of these programs are described following:
City of Portland
The City of Portland adopted an Energy Conservation Policy on August 15,
1979. Energy planners estimate that successful implementation of all components
of the plan will result in a city-wide energy savings potential of 25-35% of the
present level of energy consumption by 1995. The most controversial and pro-
gressive feature of the plan is the requirement that residences must be insulated
to cost effective levels before than can be resold, starting in 1984, and subject
to voter approval of this requirement.
B-7
-------
The plan contains six major policies, one of which is the retrofit
program for the residential sector. This program establishes a "one-stop"
conservation center which conducts an aggressive energy conservation marketing
program entitled the Energy Savings Center. The center is operated by a private
non-profit corporation that acts as a clearinghouse, coordinating information
and promoting participation in weatherization services to individuals residing
within the city limits.
In a typical Portland home, space heating consumes about 72% of the total
energy used, and a reduction in heating cequirements of 60% can be achieved if
the structure is insulated and weatherproofed. The initial stages of the retro-
fit program are optional and are accomplished through voluntary participation;
mandatory requirements for cost effective* weatherization are scheduled to take
effect in 1984, subject to voter approval. However, meeting the weatherization
requirements will then be obligatory only at the time a home is sold. The owner
will verify to the buyer than an energy audit and the necessary work has been
completed.
The City's low interest loans are currently obtainable at 8% interest
through a $3 million U.S. Department Housing and Urban Development Action Grant.
These funds are matched locally with $14 million dollars from local lending
institutions.
* Cost effectiveness standards have been determined to ensure that the cost of
weatherization combined with monthly fuel bills is no more than what was
previously paid for fuel alone.
B-8
-------
City of Seattle
Seattle City Light formed Seattle Home Insulation Program Office to
coordinate weatherization, energy audit, and financing information. The Seattle
program affects residential structures of 1-4 units. It emphasizes electrical
energy conservation but affects homes with other heat sources as well. Seattle
estimates average costs of about $1200 per home. Zero interest loans are pro-
vided by Seattle City Light for electrically-heated homes with low interest
loans for homes heated by other fuels.
Seattle requires attic ventilation_and R-30 attic insulation, ground vapor
barrier and R-19 floor insulation, waterheater jacket and thermostat setback
(130 degrees F), and insulated ducts and pipes in unheated spaces. Seattle
encourages caulking and weatherstripping, wall insulation and thermal or storm
windows. Seattle will review the progress of the voluntary weatherization pro-
gram in 1983 to determine if the mandatory program is necessary.
Fireplace Modification
, Fireplaces without glass doors or other coyering mechanism can have
negative efficiency because large quantities of heated air can escape up the
flue during nighttime hours. For example, if a fireplace fire burns between
5:00 P.M. and 11:00 P.M., the chimney damper typically cannot be shut until
about 3:00 A.M. or smoke from the dying fire will be forced into the dwelling.
Thus, for a fireplace without glass doors or similar device, heated household
air will escape up the flue all night long while little heat is produced by
the smoldering logs.
B-9
-------
For fireplaces without glass doors, negative efficiencies are often cited
because of this nighttime heat loss phenomenon. The addition of a fireplace
face-covering system can increase efficiencies to the 10 to 20% range.
Localized Room Heating
Most dwellings in the Pacific Northwest are heated by central heating
units. When room heaters are used, individuals can stay comfortable in the
heated room without the necessity of heating the air volume of the entire
dwelling. This practice is commonly employed with wood stoves where a resident
is willing to accept cooler temperatures in portions of the dwelling if the
rooms most commonly used are maintained at comfortable temperatures.
Other equipment, however, can be used to achieve similar effects by
applying heat selectively in a household. For example, the use of electric
blankets can make cooler bedroom temperatures much more acceptable, and the use
of electric heaters or heat lamps in bathrooms can reduce the need to maintain
an entire dwelling at 68 degrees to 72 degrees F temperature.
A phenomenon similar to the expansion in wood stove purchases is now
occurring with kerosene room heaters. These room heaters are commonly placed
in main living areas and maintain comfort there without the entire house being
heated to those levels. A February 1, 1982 Time magazine article states that
about 3 million households will use these kerosene room heaters in winter 1981-
1982 and that 8 to 10 million households are expected to utilize these units by
1985. Concern about safety of unvented kerosene heaters may limit the wide-
spread use of these units. Such units are illegal in Oregon under the Oregon
Uniform Mechanical Code, State Fire Marshal.
B-10
-------
Humidlfication
Humidification of the air inside a dwelling can make cooler temperatures
feel much more acceptable. Heating efficiency can be added to by increasing
the humidity inside the home. Cold air has very little water retention cap-
ability, so during winter, warmed air> is very low in relative humidity, often
under 15%. For comfort, some method of adding moisture to home air can help
significantly. A house with a 68 degree temperature and a relative humidity
of 15-20% may feel unpleasantly cool, even without drafts. Raising the relative
humidity to 35% may make the same temperature feel comfortable.
The most efficient way to humidify a wood-heated home is with the small
electric powered humidifiers. The basic units typically sell for about $100.
Humidifying is a simple job for wood stoves with flat tops. One or two sauce-
pans two-thirds full of water can be placed on top of a stove.
Relevant Government Policies
Government policies will influence the future level of residential wood
usage by actions that make wood supplies either more or less available or
expensive, by actions that effect the cost of alternate residential fuels and
by actions that affect household heating requirements. For organizational
purposes, government policies are discussed in the following categories:
Forestry Agency Policies
Policies Influencing Household Heating Costs
Forestry Agency Policies
Allocation of Forest Resources for Firewood
The bulk of this discussion focusses on Federal agency policies since
these agencies manage the vast majority of all forested government lands. For
B-ll
-------
example, in Oregon, US Forest Service and Bureau of Land Management lands
account for 84% of the current harvest on public lands
The Multiple-Use, Sustained-Yield Act of June 12, 1968 established that
the national forests shall be administered in a manner meeting a variety of
objectives, including outdoor recreation, range, timber, watershed, and wild-
life and fish purposes. No specific weighting of these different purposes is
directed. Instead, the Secretary of Agriculture is "directed to develop and
administer the renewable surface resources of the national forests for multiple
use and sustained yield of the several products and services obtained there-
from". Thus, there is a fair amount of flexibility allowed in the US Forest
Service policies.
The particular policy issue which may become an important influence on
future residential wood usage is whether government agencies will allow free or
low-cost cutting of firewood on Federal lands or whether these amounts of wood
will ultimately go to other uses such as in industrial or utility boilers or
for fiber or forest products, which can also utilize logging residues. Although
there is no stated US Forest Service policy to indicate how the US Forest Service
will allocate residues among the potential users, several US Forest Service
employees contacted in the course of this study noted that historically there
is a tendency to allocate the forest resources towards products rather than
for fuel purposes.
The precedent for allowing free cutting of firewood on US Forest Service
lands is found in the Forest Service Management Act Part 223.1 (Authorization
for Sale and Disposal of Timber) part (3) . Free use of firewood is authorized
for certain individuals and "supervisors" are granted authority to designate
portions of forest lands for personal use or domestic reasons.
B-12
-------
The management of US Forest Service lands for energy purposes is addressed
in Chapter 2170 (Energy, Management), amended in April 1980. Under Section
2170.24, the US Forest Service is directed to:
"Improve the Nation's energy alternatives by facilitating
recovery of critical fuels from forest lands and imple-
menting programs to support production and use of alter-
native fuels.1'
Responsibilities for planning and implementation activities are delineated,
but again, there is no specific directive on how much of the forest resources
should be allocated for energy programs or how wood products for energy purposes
should be allocated among potential residential, commercial, or industrial users.
In conversations with US Forest Service employees, they indicate that
future allocation of wood resources will most likely be dependent on the market
value of wood for different purposes. Several changes can be foreseen which may
increase the demand for logging residues for purposes other than for residential
wood combustion. These factors include 1) the trend toward more complete
utilization of trees by products firms; and 2) the potentially greater value of
forest products for energy conservation when used as products rather than fuel .
Up to 1974, forest product firms remo'ving timber from US Forest Service
lands were required to remove the portion of trees only over eight inches in
diameter. The tops of trees and branches with lesser diameter were typically
left behind as logging residues and burned as slash. The policy changed this
requirement to six inches diameter and a further reduction to requiring removal
of logs down to four inches diameter is scheduled to take effect in 1985. As
less and less of individual trees is left behind in forests, less wood may be
available for individuals to cut for firewood purposes.
B-13
-------
Several references in the literature note a greater value of wood products
for energy conservation when used as products rather than as fuels. For example,
Dr. F. Bryan Clark, US Forest Service Associate Deputy Chief for Research states
it takes 8 to 9 times as much energy to produce a ton of steel as a ton of wood
44
products . David M. Smith, Professor of Silviculture at Yale has noted, "The
most important way of using wood for energy conservation is as a structural
material. While it is both valuable and important to use more wood as fuel, this
52
is really only the second-best way of employing wood to save energy" . Thus,
government agencies may opt to preferentially allocate forest raw materials to
purposes other than as fuel over the next 20 years.
Counterbalancing these forces may be individuals wanting to continue to
burn wood for residential heating because of its aesthetically pleasing qualities
and heating value. The following extract from an August 1979 editorial in the
Salem (Oregon) Capital Journal is illustrative. The editorial closes:
"The Forest Service has seen enough evidence by now that
poeple want all the firewood they can get. It is up to
the agency to meet those needs in the most efficient
manner possible." 53
Allowable Time Period of Cutting
Government agencies typically allow firewood cutting on forest lands
between April through June and September through November. Many individuals
attempt to burn wood that has been cut in September through November during
the winter season immediately following. This is generally an insufficient
period to "season" or dry to proper moisture levels.
B-14
-------
Policies Influencing Household Heating Costs
The major factor influencing households to use more wood for space heating
is the potential savings achievable as compared to heating with conventional
fuels . Accordingly, government policies that effect conventional fuel prices
for residential heating can affect the relative fuel costs,therefore the
proclivity to use more and more wood for heating.
Inverted electricity rate structures have been adopted by many utilities
in the Pacific Northwest in recent years. If federal agencies were to adopt
legislation reducing the costs of conventional fuels purchased for residential
heating, there would be less incentive for using wood. However, the current
trend towards lesser federal regulation and greater reliance on free market
forces would indicate that such actions are not likely to occur in the near future
A second means by which households heating costs can be reduced is by
vigorous weatherization of dwelling structures. Weatherization can reduce a
household's heating requirement as much as 60%. Thus, future government
policies to promote weatherization such as tax credits, grants, and low interest
loans have the potential to dramatically cut average heating costs, which would
result directly or indirectly in less wood consumption.
B-15
------- |