Harvested Wood Products:
Determination of Half Lives for Wood Products in Use
Paul McFarlane1, Rob Sianchuk1, Yu Li1 and Caren Dymond2
1 Department of Wood Science,
2 BC Ministry of Forests, Lands
Faculty of Forestry,
and Natural Resource Operations
University of British Columbia.
Email: [email protected]
The forestry sector contributes to three principal carbon pools:
• Forest carbon;
• Wood products in use and
• Wood products in landfills.
Wood products
Wood products
in landfills
in use
Figure 1: Forestry sector carbon pools
Forest
- Methods -
- Results and Discussion -
The US Census Bureau has reported biennial data on housing stock, by decade of construction, from 1985-2009 (USCB 2012). A first order decay model was
fitted to each decade’s data by linear regression and the half life was then calculated using equation 1 (IPCC 2006):
half life = (ln 2)/k
- Equation 1
where: k = first order decay coefficient (years-1)
These half lives were referred to as raw data (Figure 2). Because the data only covered a period of 120 years and estimates of over 300 years were required, a
smoothed regression model was developed by log transformation to enable the long term calculation of half lives. This log regression model was then used to
produce a decay curve for US single family homes (Figure 3). An optimal first order decay curve, which is the IPCC default approach (IPCC 2006), was then
fitted to these data using a sum of least squares approach and the first order decay half life for this curve was determined. A Gamma distribution, which has
been suggested as being a better approach than first order decay (Marland et al 2010), was also fitted to the data using a sum of least squares method. The
half life for this data set was also determined.
The half lives of single family homes in the US varied substantially by decade of
construction. Houses built in the 1990’s had a half life of 352 years while those
built before 1920 had one of 74 years (Figure 2). This distribution of half lives is
Due to a lack of reliable fundamental information, most HWP
half-lives have been based on best guesses rather than
analysis of hard data, (Pingoud et al 2003). Previously
published HWP half-life estimates for housing are summarized
in Table 1. Estimates range from the IPPC default value of 30
years to Skog and Nicholson’s assessment of 100 years.
Log regression model
0.9
350
300
Raw data
250
Half life (years)
The lifetimes of these materials vary from short periods in the
case of bioenergy or paper to long times for wood used in
houses. The United States (US) housing market has historically
been the major consumer of Canadian wood products and it is
therefore important to quantify this HWP pool. In order to do
this, the rate at which houses are removed from use needs to
be determined and the half life is the parameter used by the
IPCC to determine housing removal rates (IPCC 2006).
1
400
Although most of the focus has been on quantifying forest
carbon, wood products have the potential to sequester
significant quantities of carbon (Bowyer et al 2010). These
materials are referred to as harvested wood products (HWP)
and they are defined as “wood-based materials that, following
harvest, are transformed into commodities such as furniture,
plywood, paper and paper-like products or used for energy”
(FCCC 2003).
Log regression model
200
150
100
50
Fraction of single family homes remaining in stock
- Introduction -
IPCC default
0.8
0.7
Optimized 1st order decay to log regression
model
0.6
Half life = 101 years
0.5
0.4
0.3
- Conclusions -
0.1
0
1920s
1930s
1940s
1950s
1960s
1970s
1980s
0
1990s
20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360
Time in use (years)
Decade of construction
Figure 4. First order decay curves – IPCC default and optimal fit to decay curve
Figure 2. Half life data for US single family homes by decade of construction
1
This study addresses this issue for single family homes in the
United States and, to our knowledge, it is the first detailed
regression analysis published on housing half lives.
Solid wood products
Half life estimate
(years)
Reference
30
IPCC 2006
0.9
First order decay regression model
Fraction of single family homes remaining in stock
Product category
(IPCC default)
Structural building
50
EFI 2002
materials
Single family homes
80
(pre-1980)
Single family homes
(post-1980)
Skog & Nicholson
1998
100
Skog & Nicholson
of the housing decay rates in the US and will substantially underestimate the
carbon mass in this HWP pool.
• The Gamma distribution gave an adequate fit to the data and the half life was
estimated to be 113 years. It is recommended that the Gamma distribution is
used to model pools of HWP’s in single family homes.
IPCC first order decay default
- Acknowledgements and References-
Gamma distribution alpha= 1.8 beta =76
This research was funded by the Pacific Institute for Climate Solutions
0.7
0.6
Half life = 113 years
0.5
0.4
0.3
0.2
1998
0
0
20
40
60
80
100 120 140 160 180 200 220 240 260 280 300 320 340 360
Time in use (years)
Figure 3. Decay curve for US single family homes
• The half lives of US single family homes vary substantially with the decade of
their construction, with older houses having shorter half lives.
• The half lives were substantially greater than IPPC default values.
• The IPCC default methodology and half live value provided a poor estimate
0.8
0.1
Table 1: Published estimates of wood product half lives
Consequently, the IPPC default methodology and half life value provide an
inadequate estimate of the decay pattern of US single family homes and this
methodology will substantially underestimate the mass of HWP in this pool.
The Gamma distribution provided a better fit to the sigmoidal shape of the US
housing decay curve (Figure 5). The best fit was obtained by a Gamma
distribution with alpha = 1.8 and beta = 76, which and a half life of 113 years and it
is recommended that this methodology is adopted for quantifying this HWP pool.
0.2
0
< 1920
intuitively rational, with older houses having a greater probability of demolition
than newer houses. The IPCC default methodology does not take into account
this variation in half lives with time as the first order decay method uses a
constant half life irrespective of the construction period.
The log regression approach provided an adequate model of the half life data but
underestimated the values for older and younger homes (Figure 2). The US single
family home decay curve created using this model fitted the raw data well (Figure
3). In contrast, the IPCC default half life and model substantially underestimated
the data, as shown by the pale blue line on Figure 4. The first order decay model
that was optimally fitted using a sum of least squares approach provided a better
estimate but had difficulty fitting the inverse sigmoidal shape of the housing decay
curve (Figure 4). The half life determined by this model was 101 years.
Figure 5. Optimal fit of Gamma distribution to decay curve
Bowyer , J., Bratkovich, S. Howe, J., Fernholz, K. 2010: Recognition of carbon storage in harvested wood
products: post- Copenhagen update. Dovetail Partners report. February 25. Minneapolis, USA. 19 pp
EFI 2002: The impacts of manufacturing and utilisation of wood products
on the European carbon budget. European Forest Institute Internal Report 9, 2002; Joensuu, Finland.
http://www.efi.fi/publications/Internal Reports/IR 9.pdf.
FCCC 2003. Estimation, reporting and accounting of harvested wood products. United Nations Framework
Convention on Climate Change Technical Paper FCCC/TP/2003/7. 44pp.
IPCC 2006: Harvested wood products, Volume 4, Chapter 12. in IPCC Guidelines for National Greenhouse
Gas Inventories, Prepared by the National Greenhouse Gas Inventories Programme, Eggleston H.S.,
Buendia L., Miwa K., Ngara T. and Tanabe K. (eds). 33pp. Marland, E.S., Stellar, K. and Marland, G.H.
2010. A distributed approach to accounting for carbon in wood products. Mitig Adapt Strat Glob Change 15:
71-91
Pingoud, K., Perälä, A.-L., Soimakallio, S. & Pussinen, A. 2003. Greenhouse gas impacts of harvested
wood products - Evaluation and development of methods. VTT Technical Research Centre of Finland. 120
pp. ISBN 951.38.6189.9 (URL: http://www.vtt.fi/inf/pdf/).
Skog, K. and Nicholson, G. 1998. Carbon cycling through wood products: The role of wood and paper
products in carbon sequestration. Forest Products Journal 48(7/8), 75–83.
USCB 2012. American Housing Survey. http://www.census.gov/hhes/www/housing/ahs/nationaldata.html