Analytical framework
BC’s forests cover about 55 million hectares, of which 95% are owned by the provincial government. BC is the largest exporter of softwood lumber in the world and the largest bioenergy producer in North America. It is estimated that about 6–7 billion tonnes of carbon are stored in the aboveground biomass (95% of which are certified by third-party certification) (BCMoFLNRO 2013), with an average net carbon removal from the atmosphere of 62.8 MtCO2e/year during the last 25 years (Government of British Columbia 2016), which is equivalent to the total annual CO2e emissions from all other sectors in BC. Harvesting of BC’s forests transfers roughly the same amount of carbon (66.2 MtCO2e/year) to wood products. These characteristics suggest that the mitigation potential in BC’s forest sector could be substantial if appropriate strategies are implemented.
In line with the IPCC’s definition of mitigation (IPCC 2014), our analysis considered mitigation potential as reduced greenhouse gas (GHG) emission or enhanced carbon sequestration that would result from implementation of a mitigation option, relative to a baseline. Such an approach canceled out all factors that were assumed to be unchanged between the baseline and a mitigation scenario, including variables with uncertainties like GHG emissions from wildfires. A systems perspective was employed to include potential mitigation resulting from changes in forest management, use of longer-lived products (LLPs) and bioenergy, as well as avoided emissions in other sectors due to displacement effects (Fig. 1). We defined forest sector mitigation based on carbon stock changes in BC’s forest ecosystems and in harvested wood products manufactured from wood that was harvested in BC regardless of where in the world these products reside—the IPCC production approach for estimation of HWP C balances (IPCC 2013). Domestic mitigation is the sum of forest sector mitigation plus displacement effects in BC resulting from the use of BC harvested wood products. Global mitigation is the sum of domestic mitigation plus the displacement effects that occur outside BC as a consequence of the use of HWP manufactured from wood harvested in BC. All displacement factors assumed that concrete and steel would be used as an alternative to BC wood, rather than wood from elsewhere. At all scales, we did not consider possible leakage effects due to, for example, displacement from imported wood products, or imperfect substitution due to market interactions—we focused on the mitigation that BC’s forest sector may contribute rather than the net mitigation benefit for the global atmosphere. The exclusion of leakage effects in this analysis is justified by the scale of our analysis. Leakages from shifting harvest and land-use change need to be considered at the scale of individual offset projects but are expected to be minimal at the provincial scale, because changes resulting from mitigation strategies were small compared to the size of BC’s forest sector.
Our analysis was conducted at a spatial resolution with 74 spatial units based on the forest management units (FMUs) identified in Canada’s 2014 National GHG Inventory Report (Environment Canada 2016). These FMUs in BC were defined by the boundaries of Timber Supply Areas (TSA) and Tree Farm Licences (TFL) and categorized in five ecozones and three forest regions (Fig. 2). The mitigation potential in each FMU was examined for the period from 2017 to 2050, within which three periods were of particular interest in terms of BC’s and Canada’s GHG emission reduction targets: 2017–2020 (short term), 2017–2030 (medium term), and 2017–2050 (long term).
In this study, eleven different strategies were assessed relative to the baseline, including five individual strategies and six combinations of the individual strategies (Table 1). The mitigation strategies were developed in consultation with experts and professionals from the BC Ministry of Forests, Lands, and Natural Resource Operations (BCMoFLNRO), the BC Ministry of Environment, and other organizations. Although some of the strategies are nominally similar to those in the national study of Smyth et al. (2014), all strategies in this study were adjusted based on the biophysical and economic characteristics of BC’s forest sector, as well as the province’s mitigation needs (see “Appendix” for further details of the five individual strategies). Some strategies were implemented with a ramp-up period of 2017–2020 as it was expected that time would be needed to scale-up efforts.
Table 1 Individual mitigation strategies and their combinations
The baseline was defined as the forest management activities and use of HWP that would occur in the absence of mitigation activities. Modeling assumptions for the historical period (1990–2012) in the baseline were based on Canada’s GHG National Inventory Report (Environment Canada 2016). We included harvest and wildfire projections for each FMU for the future time period (2013–2050) based on a forecast of future harvest levels and historical average annual area burned during 1990–2012, respectively.
Mitigation effects for each strategy are expected to vary across FMUs depending on various factors, such as size of the spatial unit, forest characteristics, harvest levels, potential for substitution of fossil fuels, and logging and transportation costs. Implementation of the strategies was modeled for each FMU independently, i.e., there were no interactions between FMUs. We recognized that there is no “best-for-all” strategy for the province; rather, a portfolio of strategies with FMU-specific strategy selections would generate greater mitigation and/or require lower cost. Therefore, we first estimated attributes of interest (e.g., global, domestic or forest sector mitigation, or mitigation cost) for all strategies in each FMU and then constructed a specific portfolio by selecting the strategy to best meet the portfolio goal in each FMU over a specific time period. Among others, four portfolios for the long term (2017–2050) were of particular interest: (1) portfolio that maximizes global mitigation (PORT1), (2) portfolio that maximizes domestic mitigation (PORT2), (3) portfolio that maximizes forest sector mitigation (PORT3), and (4) portfolio that minimizes domestic mitigation cost (PORT4) (see Fig. 5 for all portfolios).
Mitigation impact of strategies
Forest ecosystem carbon dynamics in this study were estimated using the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3) (Kurz et al. 2009) with historical datasets for BC from the National Forest Carbon Monitoring, Accounting and Reporting System (Kurz and Apps 2006; Stinson et al. 2011). Carbon transferred from forest ecosystems to HWP and bioenergy were tracked through manufacturing, use/export, and end-of-life use by the Carbon Budget Modeling Framework for Harvested Wood Products (Smyth et al. 2014). More details about the carbon models are described in the “Appendix”.
Following Smyth et al. (2014), we defined mitigation as the difference in GHG emissions between the baseline and a scenario with mitigation actions:
$$ M={E}_{\mathrm{base}}-{E}_{\mathrm{strategy}} $$
where M is the mitigation and E
base and E
strategy are the net emissions in the baseline scenario and the mitigation scenario, respectively. In both scenarios, net emissions included three components—the forest ecosystem emissions, the HWP emissions including bioenergy, and emissions as a result of displacement. Displacement impacts for products included the emissions associated with the extraction and transportation of raw materials and the manufacturing of products, based on aggregated studies in the literature (Smyth et al. 2016).
To obtain the total cumulative global mitigation impact of a strategy, we aggregated FMU results for the strategy to the provincial level. Emissions associated with exported HWPs were taken into account in order to be consistent with the IPCC Production Approach in accordance with internationally agreed guidance (IPCC 2013). Displacement effects that occurred abroad were included in the global mitigation impact of a strategy but not in the domestic mitigation impact which was used for economic analyses, as mitigation resulting from displacement occurring abroad is not included in Canada’s national or provincial GHG inventories (Environment Canada 2016). We also assumed that bioenergy from harvest residues was produced and consumed only within BC.
Two displacement effects were considered in all individual strategies and associated combinations involving a change in HWP or bioenergy production: substitution between solid wood products (panels) and emission-intensive materials (concrete and steel) in housing construction and substitution between bioenergy from harvest residues and fossil fuel energy in providing power and heat. Emissions related to displacement were calculated by employing displacement factors for sawnwood, panels, and bioenergy. Displacement factors are commonly used to indicate how many tonnes of carbon emissions from alternatives can be avoided per tonne of carbon in wood-based products used (Sathre and O’Connor 2010). In this study, we applied national average displacement factors of 2.1 tC/tC for sawnwood and 2.2 tC/tC for panels derived from Smyth et al. (2016) for the entire period by considering three end-use products: single-family home, multi-family home, and multi-use building. For avoided emissions for bioenergy in BC, we estimated displacement factors using a linear programming (LP) model that maximized avoided emissions in each FMU by selecting different bioenergy facilities to substitute for the most emission-intensive fuel sources that would have been used for baseline electricity and heat generation. Smyth et al. (2016) provided a detailed description of the LP model, and details of the bioenergy facilities are summarized in Table 2.
Table 2 Bioenergy facility types and characteristics adapted from Smyth et al. (2016)
Mitigation costs of strategies
The domestic mitigation costs (Canadian dollars) were estimated using the Model for Economic Analysis of Forest Carbon Management (MEA-FCM) which was originally designed and employed by Lemprière et al. (2017). We only considered domestic mitigation rather than global mitigation, because mitigation costs associated with BC’s forest sector and related local industries were of interest in this study. We defined mitigation cost (TC) as the total cost to implement a mitigation strategy which equals the change between the baseline and a mitigation scenario in the present values of the total net revenues (NR) of both the forest sector and the other industries/sectors affected by displacement:
$$ TC=\varDelta {NR}_{\mathrm{forest}}+\varDelta {NR}_{\mathrm{dis}} $$
where ΔNR
forest is the total net revenue change in the forest sector, and ΔNR
dis is the total net revenue change in other industries/sectors. The total net revenues of the forest sector (NR
forest) in either the baseline or a mitigation scenario can be further broken down as follows:
$$ {NR}_{\mathrm{forest}}={R}_{\mathrm{fm}}-{C}_{\mathrm{fm}}+{R}_{\mathrm{hwp}}-{C}_{\mathrm{hwp}} $$
where R
fm refers to total revenue in forest management via harvesting, and R
hwp refers to total revenue from wood product manufacturing and bioenergy production; and C
fm and C
hwp are the associated total costs. The total revenue change in the forest sector was calculated by taking the differences in all components between the baseline and a mitigation scenario.
For the total net revenue change in other industries/sectors, we considered concrete and steel industries and part of the energy sector that generates electricity and heat using fossil fuels:
$$ \varDelta {NR}_{\mathrm{dis}}=\left({p}_{\mathrm{c}}-{c}_{\mathrm{c}}\right)\times \varDelta {Q}_{\mathrm{panel}}\times {u}_{\mathrm{c}}+\left({p}_{\mathrm{s}}-{c}_{\mathrm{s}}\right)\times \varDelta {Q}_{\mathrm{panel}}\times {u}_{\mathrm{s}}+\left({p}_{\mathrm{e}}-{c}_{\mathrm{e}}\right)\times \varDelta {Q}_{\mathrm{residue}}\times {u}_{\mathrm{e}} $$
where p
c, p
s, and p
e and c
c, c
s, and c
e refer to the per unit prices and costs of concrete and steel products, as well as fossil fuel energy, respectively; u
c and u
s are parameters that indicate how many tonnes of concrete and steel can be substituted per cubic meter of panels, respectively, and u
e is the parameter that indicates the amount of bioenergy (MWh) that can be produced per cubic meter of captured harvest residues. Bioenergy displaced power and heat for residential and industrial uses, and for simplicity, we measured both energy types using MWh. ΔQ
panel and ΔQ
residue are the volume changes in panel production and harvest residues between the baseline and the strategies. Note that u
c and u
s are constants while u
e varied across FMUs as each had a different fuel mix that was displaced by harvest residues. To be consistent with the calculation of displacement factors, we assumed that all concrete and steel products are domestically produced in BC—we included net revenue changes in those two industries in the mitigation cost, though we realized that most steel used in BC is imported and thus the profit changes in the steel industry actually occur outside of BC.
The cost per tonne ($/tCO2e) of domestic mitigation in each FMU for each strategy was then calculated by dividing mitigation cost by domestic mitigation impact over the time period:
$$ {MC}_{ni}=\frac{PTC_{ni}}{PDE{}_{ni}} $$
where MC
ni
is the cost per tonne in FMU i for strategy n, PTC
ni
is the present value (2016 as the base year) of the total mitigation cost during 2017–2050 with a 3% discount rate, and PDE
ni
is the present value of the change in total domestic emissions between the baseline and the strategy scenario with a 1% discount rate. The 3% real social discount rate was selected based on the Canadian Cost-Benefit Analysis Guide (TBS 2007). The 1% discount rate for carbon emissions was derived from the social discount rate adjusted for the marginal social cost of damages resulting from emissions—we assumed that physical carbon emissions are a proxy for the social cost of the damages, and the marginal damage of emissions will grow at an annual rate of 2% (Greenstone et al. 2013). Therefore, a 3% social discount rate for monetary value and a 2% rate for marginal damage of emissions imply a 1% discount rate for carbon emissions. By discounting both mitigation cost and quantity, cost per tonne of mitigation was considered a measure of cost effectiveness of a mitigation strategy over the entire period of 2017–2050. We kept the 2% difference between the discount rates for monetary values and the carbon emissions in the sensitivity analysis to keep the marginal damage of emission unchanged over time.
For the prices and costs for harvesting and products, we used annual averages to reflect long-term trends and assumed that they did not change over time. Detailed assumptions were developed in consultation with BCMoFLNRO and FPInnovations and summarized in Tables 3, 4, 5, and 6. Softwood/hardwood log costs in the baseline were estimates derived from log cost surveys. The log cost includes tree-to-truck cost, hauling cost, cost of stumpage, and costs for forest planning and administration, road development and management, and silviculture. Prices and costs for salvage logging and other industrial roundwood were assumed to be the same as those for regular harvest and sawnwood, respectively. The higher utilization strategy was assumed to slightly decrease the log cost per cubic meter, because the harvest volume was kept unchanged but the harvested area was smaller. The harvest less strategy reduced the harvest volume and was assumed to increase the logging cost since cut blocks were assumed to be more dispersed in order to match the same timber characteristics as in the baseline. A similar assumption was made for the restricted harvest strategy as more young stands would be harvested to meet the target harvest level. No price/cost changes in harvesting were assumed in the harvest residue for bioenergy (hereafter bioenergy) strategy.
Table 3 Harvest cost and price assumptions for individual strategies ($/m3 in 2014 dollars)
Table 4 Price assumptions for individual strategies for harvested wood products and bioenergy (2014 dollars)
Table 5 Cost assumptions for individual strategies for harvested wood products and energy (2014 dollars)
Table 6 Cost and price assumptions for substituted products in strategies involving displacement effects (2014 dollars)
Harvests were used to produce generic HWP commodities: sawnwood, other industrial roundwood, panels, and pulp and paper products (Tables 4 and 5). In the more LLP strategy, we assumed a 2% increase in the pulp and paper manufacturing cost and a 2% decrease in the panel production cost due to economies of scale in existing mills. Prices and costs for bioenergy were determined by the LP model for each FMU. Since we assumed there was no bioenergy production from harvest residues in the baseline, no baseline prices and costs were needed. In the bioenergy strategy, the price of bioenergy in each FMU was calculated based on generic electricity price ($120/MWh) and heat price ($8/GJ) weighted by proportions of power and heat that bioenergy generated. The bioenergy production cost in the bioenergy strategy was estimated by the LP model based on the selected facilities and associated production costs (Table 2). The bioenergy production cost also included costs for processing harvest residues and transporting to facilities, as well as the avoided cost of reduced slashburning (Table 7). No cost was assumed for extracting harvest residue from cut blocks to roadside, because the full-tree harvesting approach was assumed to be employed in BC.
Table 11 Average annual domestic mitigation and associated economic and socio-economic impacts (in 2016 Canadian dollars), 2017–2050
Economic assumptions for displacement included prices and costs for concrete and steel products and fossil fuel energy. The prices for fossil fuel energy in the bioenergy strategy were the same as the prices for bioenergy, while the unit costs were calculated by dividing the total energy production cost (including fuel cost and production cost) from all fuel sources being substituted by the total bioenergy production (Table 8).
Table 7 Cost assumptions for the supply of harvest residues (in 2008 dollars)
Socio-economic impacts of strategies
The socio-economic impacts of mitigation strategies were analyzed using multipliers from the national input-output (I/O) model (Statistics Canada 2014). The value of a multiplier refers to the increase/decrease in an indicator (e.g., gross domestic product (GDP)) if the demand for the output of a given industry increases/decreases by $1 (or $1 million for employment). In this study, multipliers were used to assess impacts on employment, GDP, and government revenues in BC’s economy in response to changes in the forest sector resulting from the implementation of mitigation strategies. We considered both direct effects and indirect effects on those indicators, where the former refers to the impacts directly induced from a change in an industry’s output, and the latter measures the impacts of further output changes due to interactions among industries within BC in response to the initial changes in the directly affected industry. Given that induced effects may cause double counting (Horne 2008), we did not consider induced effects in our analysis. We also only focused on the socio-economic impacts in response to changes in BC’s forest sector—no impacts resulting from displacement effects in other industries/sectors were estimated.
Five different industries in the forest sector as defined in the North American Industry Classification System (NAICS) (Table 9) were used to estimate socio-economic impacts. Mitigation actions with harvest-related activities were assigned to “forestry and logging”; strategies involving HWP were linked to two manufacturing industries—“wood products manufacture” and “pulp, paper, and paperboard mills”; for the bioenergy strategy, multipliers from the “electric power generation, transmission and distribution” were employed for bioenergy generation, and averages of multipliers for forestry and logging and “truck transportation” were used to represent impacts of harvest residue extraction for bioenergy since this activity is not specified in NAICS. For strategies involving multiple industries, total impacts were estimated as the sum of impacts on all of the relevant industries.
Table 8 Cost assumptions for power and heat production using fossil fuels (2008 dollars)
Our analysis estimated the direct and indirect effects using multipliers shown in Table 9. For each mitigation strategy, the initial change in an industry was estimated as the change in total revenues of that industry between the baseline and a mitigation scenario. Following the initial change, total direct/indirect socio-economic impacts on GDP (SE
GDP) for the entire period (2017–2050) were estimated using the corresponding multipliers:
$$ {SE}_{\mathrm{GDP}}=\sum_j^J\sum_t^T\frac{GDP_{j t}}{{\left(1+ r\right)}^t}=\sum_j^J\sum_t^T\frac{\Delta { T R}_{j t}\times {m}_j^{GDP}}{{\left(1+ r\right)}^t} $$
where GDP
jt
is the GDP impact from industry j at time t, ΔTR
jt
is the total revenue change estimated in MEA-FCM and \( {m}_j^{GDP} \) is the multiplier for direct/indirect GDP impacts of industry i, and r is the discount rate.
Because there are no multipliers specified for impacts on government revenue in the national I/O model, we used the sum of the multipliers for the government revenue-related components that are used for GDP calculations (personal and business income taxes were not included). We did not separately estimate impacts on provincial government revenue in BC since no associated multipliers were available; rather, we estimated the impact on total government revenue for all federal, provincial, and municipal governments:
$$ {SE}_{\mathrm{revenue}}=\sum_j^J\sum_t^T\frac{GR_{j t}}{{\left(1+ r\right)}^t}=\sum_j^J\sum_t^T\frac{\Delta { T R}_{j t}\times \left({m}_j^{T p}+{m}_j^{T n}-{m}_j^{Sp}-{m}_j^{Sn}\right)}{{\left(1+ r\right)}^t} $$
where GR
jt
is the total government revenue impact from industry j at time t; \( {m}_j^{Tp} \) and \( {m}_j^{Tn} \) are multipliers for taxes on products and production of industry j, respectively; and \( {m}_j^{Sp} \) and \( {m}_j^{Sn} \) are multipliers for subsidies on products and production of industry j, respectively.
The direct impacts on employment were not examined using employment multipliers because most mitigation actions involved in the strategies were so specific that the available multipliers were too general and cannot reflect their impacts on employment appropriately. Instead, we estimated the direct impacts on jobs (\( {SE}_{\mathrm{direct}}^{\mathrm{Job}} \)) by multiplying a labor intensity parameter (e.g., person-year (PY)/m3, see Table 10) for each industry with the corresponding biophysical change that occurred in that industry when a given mitigation strategy was implemented:
Table 9 Industries identified for the forest (including bioenergy) sector and associated multipliers
$$ {SE}_{\mathrm{direct}}^{\mathrm{Job}}=\sum_j^J\sum_t^T{PY}_{j t}=\sum_j^J\sum_t^T\Delta {Q}_{j t}\times {k}_j $$
where PY
jt
refers to the direct change in person-year (the amount of work done by one person in a year) required in industry j at time t, ΔQ
jt
represents the biophysical changes (e.g., harvest volume, volume of harvest residue for bioenergy, or HWP or bioenergy production) due to strategy implementation, and k
j
is the labor intensity parameter for industry j.
Specifically, the labor intensity required for every cubic meter of fiber harvested or manufactured was estimated for each of the forestry and logging, wood products manufacture, and pulp, paper, and paperboard mills’ industries by dividing the annual average of the total number of employees in each industry by the annual average of total industrial roundwood harvest in BC during 2009–2013. The data were derived from Statistics Canada (CANSIM 2015) and the National Forestry Database (NFD 2015), respectively. We used estimates in FPInnovations (2010) as the labor intensity for harvest residue extraction, and the labor requirement for an 8 MWe combined heat and power (CHP) steam turbine (FPAC and FPInnovations 2011) to estimate the labor intensity for bioenergy generation.
Although we did not use multipliers to calculate the direct effects on employment, we assumed that the ratio between multipliers for direct and indirect effects on employment is accurate. We then used the ratio and the estimated direct effects described previously to calculate the indirect impacts on jobs (\( {SE}_{\mathrm{indirect}}^{\mathrm{Job}} \)):
$$ {SE}_{\mathrm{indirect}}^{\mathrm{Job}}=\sum_j^J\sum_t^T{PY}_{j t}^{\hbox{'}}=\sum_j^J\sum_t^T{PY}_{j t}\times \frac{m_j^{\hbox{'}}}{m_j} $$
where \( {PY}_{jt}^{\hbox{'}} \) is the indirect change in PY required in industry j at time t and m
j
and \( {m}_j^{\hbox{'}} \)are multipliers for direct and indirect effects on employment, respectively.