Different carbon accounting methodologies have a significant impact on the calculated results and therefore on the scientific arguments. Generally, the purpose of carbon accounting is to determine the carbon intensity of an energy scenario, i.e. whether it is carbon positive, carbon neutral or carbon negative. Carbon positive scenarios are likely to be net emitters of CO₂, carbon negative projects are net absorbers of CO₂, while carbon neutral projects balance emissions and absorption perfectly.^[a]

As a consequence of both natural causes and human practices, carbon continually flows between carbon pools, for instance the atmospheric carbon pool, the forest carbon pool, the harvested wood products carbon pool, and the fossil fuels carbon pool. When the carbon level in pools other than the atmospheric carbon pool increase, the carbon level in the atmosphere decrease, which helps mitigate global warming.^[b] If a researcher measures the amount of carbon moving from one pool to another, they can gain insight and recommend practices that maximize the amount of carbon stored in carbon pools other than the atmospheric carbon pool. Three concepts are especially important, namely carbon debt, carbon payback time and carbon parity time:

Carbon debt accrues when biomass is removed from growing sites, for instance forests. It is counted when the trees are felled because the UNFCCC (the UN organization that countries report their emissions to) has decided that emissions should be counted already at this point in time, instead of at the combustion event.^[c]

Carbon payback time is the time it takes before this carbon is "paid back" to the forest, by having the forest re-absorb an equivalent amount of carbon from the atmosphere.

Carbon parity time is the time it takes for one energy scenario to reach carbon parity with another scenario (i.e. store the same amount of carbon as another scenario.)^[d] One of these scenarios can for instance be a bioenergy scenario, with carbon counted as stored in the part of the forest that was not harvested, and carbon counted as lost for the amount of forest that were harvested (cf. the UNFCCC rule mentioned above.) However, the amount of carbon that resides in woody construction materials and biofuels made from this harvest can be "counted back" into the bioenergy scenario's carbon pools for the amount of time it takes before this carbon decays naturally or are burnt for energy. The alternative scenario can for instance be a forest protection scenario, with carbon counted as stored in the whole forest – a forest that is bigger than in the bioenergy scenario because no trees were harvested at all, and in addition also continued to grow (while waiting for the carbon stored in the bioenergy scenario to catch up to its own carbon level.)^[e] However, the implied "lock-in" of carbon in the forest also means that this carbon no longer is available for production of woody construction materials and biofuels, which means that these have to be replaced by other sources. In most cases, the most realistic sources are fossil sources, which means that the forest protection scenario here will be "punished" by having the fossil fuel emissions it is responsible for subtracted from its carbon pool. (Note that this fossil carbon is often instead technically speaking counted as added to the bioenergy carbon pool (instead of subtracted from the no-bioenergy carbon pool), and called "displaced" or "avoided" fossil carbon.)

A net carbon debt for the bioenergy scenario is calculated when the net amount of carbon stored in the forest protection scenario's carbon pool is larger than the net amount of carbon stored in the bioenergy scenario's carbon pools. A net carbon credit for the bioenergy scenario is calculated when the net amount of carbon stored in the forest protection scenario's carbon pool is smaller than the net amount of carbon stored in the bioenergy scenario's carbon pools.^[1] The carbon parity time then is the time it takes for the bioenergy scenario to go from debt to credit.^[f]

To recap, a project or scenario can be assessed solely on its own merits, specifically the time it takes to pay back removed carbon (carbon payback time.) However, it is common to include alternative scenarios (also called "reference scenarios" or "counterfactuals") for comparison.^[g] When there is more than one scenario, carbon parity times between these scenarios can be calculated. The alternative scenarios range from scenarios with only modest changes compared to the existing project, all the way to radically different ones (i.e. forest protection or "no-bioenergy" counterfactuals.) Generally, the difference between scenarios is seen as the actual carbon mitigation potential of the scenarios.^[h] In other words, quoted emission savings are relative savings; savings relative to some alternative scenario the researcher suggest. This gives the researcher a large amount of influence over the calculated results.

In addition to the choice of alternative scenario, other choices has to be made as well. The so-called "system boundaries" determine which carbon emissions/absorptions that will be included in the actual calculation, and which that will be excluded. System boundaries include temporal, spatial, efficiency-related and economic boundaries:^[i]

The temporal boundaries define when to start and end carbon counting. Sometimes "early" events are included in the calculation, for instance carbon absorption going on in the forest before the initial harvest. Sometimes "late" events are included as well, for instance emissions caused by end-of-life activities for the infrastructure involved, e.g. demolition of factories. Since the emission and absorption of carbon related to a project or scenario changes with time, the net carbon emission can either be presented as time-dependent (for instance a curve which moves along a time axis), or as a static value; this shows average emissions calculated over a defined time period.

The time-dependent net emission curve will typically show high emissions at the beginning (if the counting starts when the biomass is harvested.) Alternatively, the starting point can be moved back to the planting event; in this case the curve can potentially move below zero (into carbon negative territory) if there is no carbon debt from land use change to pay back, and in addition more and more carbon is absorbed by the planted trees. The emission curve then spikes upward at harvest. The harvested carbon is then being distributed into other carbon pools, and the curve moves in tandem with the amount of carbon that is moved into these new pools (Y axis), and the time it takes for the carbon to move out of the pools and return to the forest via the atmosphere (X axis). As described above, the carbon payback time is the time it takes for the harvested carbon to be returned to the forest, and the carbon parity time is the time it takes for the carbon stored in two competing scenarios to reach the same level.^[j]

The static carbon emission value is produced by calculating the average annual net emission for a specific time period. The specific time period can be the expected lifetime of the infrastructure involved (typical for life cycle assessments; LCA's), policy relevant time horizons inspired by the Paris agreement (for instance remaining time until 2030, 2050 or 2100),^[2] time spans based on different global warming potentials (GWP; typically 20 or 100 years),^[k] or other time spans. In the EU, a time span of 20 years is used when quantifying the net carbon effects of a land use change.^[l] Generally in legislation, the static number approach is preferred over the dynamic, time-dependent curve approach. The number is expressed as a so-called "emission factor" (net emission per produced energy unit, for instance kg CO₂e per GJ), or even simpler as an average greenhouse gas savings percentage for specific bioenergy pathways.^[m] The EU's published greenhouse gas savings percentages for specific bioenergy pathways used in the Renewable Energy Directive (RED) and other legal documents are based on life cycle assessments (LCA's).^[n][o]

The spatial boundaries define "geographical" borders for carbon emission/absorption calculations. The two most common spatial boundaries for CO₂ absorption and emission in forests are 1.) along the edges of a particular forest stand and 2.) along the edges of a whole forest landscape, which include many forest stands of increasing age (the forest stands are harvested and replanted, one after the other, over as many years as there are stands.) A third option is the so-called increasing stand level carbon accounting method:

– In stand level carbon accounting, the researcher may count a large emission event when the stand is harvested, followed by smaller, annual absorption amounts during the accumulation phase that continues until the stand has reached a mature age and is harvested again.

– Likewise, in increasing stand level accounting, the researcher counts a large emission event when the stand is harvested, followed by absorption of smaller quantities of carbon each year during the accumulation period. However, one year after the first harvest, a new stand is harvested. The researcher do not count the carbon that was absorbed in this second stand after the first, neighbouring stand was harvested, only the large emission at the harvest event of the second stand. The next year the same procedure repeats for the third stand; the carbon that was absorbed by this stand after the harvest of the first and second stand is not counted, while a large emission amount is counted when the third stand is harvested. In other words, in increasing stand level accounting the whole carbon account is composed of a number of individual stand-level accounts, each with its own, individual starting point.

– In landscape level accounting, the researcher counts a large emission event when the first stand is harvested, followed by absorption of smaller quantities of carbon each year during the accumulation period for this particular stand. Like with increasing stand level accounting, a new stand is harvested the second and third year etc., and these emission events are all counted. Unlike with increasing stand level accounting however, the researcher also counts the carbon that is absorbed by all stands after the harvest of the first stand in the forest landscape. In other words, instead of calculating carbon emissions from a lot of different starting points, forest landscape accounting uses only one, common starting point for the whole forest landscape, namely the year the first stand was harvested.^[3]

So, the researcher has to decide whether to focus on the individual stand, an increasing number of stands, or the whole forest landscape.

According to Lamers et al., the stand level spatial boundary choice is typical for early carbon modeling, and it leads to carbon cycles that resemble sawteeth (dramatic increases in emissions at harvest, followed by slow declines as the forest stand absorb carbon.) The key benefit of stand-level analysis is its simplicity, and this is the primary reason for it still being part of today's carbon analyses. However, while the study of single stands provide easily comprehensible results (for example on the carbon effects of different harvesting choices), real-world timber/woody biomass supply areas consist of several stands of different maturity, for instance 80. Over a time period of 80 years then, all stands are successively harvested and replanted. To accurately calculate the carbon flow over such large areas the spatial boundary of the calculation has to increase from stand level to landscape level, as the forest "landscape" contains all the individual forest stands.^[4] Cowie et al. argue that landscape level accounting is more representative of the way the forestry sector manages to produce a continuous supply of wood products.^[p] The IPCC recommends landscape-level carbon accounting as well (see Short-term urgency below).

Further, the researcher has to decide whether emissions from direct/indirect land use change should be included in the calculation. Most researchers include emissions from direct land use change, for instance the emissions caused by cutting down a forest in order to start some agricultural project there instead. The inclusion of indirect land use change effects is more controversial, as they are difficult to quantify accurately.^[q][r] Other choices involve defining the likely spatial boundaries of forests in the future. For instance, is increased harvesting and perhaps even forest expansion more realistic than forest protection in a situation with high demand for forest products? Or alternatively, is smaller forests perhaps more realistic than forest protection in a situation with low demand for forest products and high demand for new land or new areas for housing and urban development? Lamers & Junginger argue that from a nature conservation and carbon strategy evaluation perspective, forest protection is a valid option. However, protection is unlikely for forest plantations – in the absence of demand for forest products (e.g., timber, pulp or pellets), "[...] options such as conversion to agriculture or urban development may be more realistic alternatives [...]."^[s] Cowie et al. argue that privately owned forests are often used to create income and therefore generally sensitive to market developments. Forest protection is an unrealistic scenario for most of the privately owned forests, unless forest owners can be compensated for their loss of income.^[t] According to the EU's Joint Research Centre, 60% of the European forests are privately owned.^[5] In the US, over 80% is privately owned in the east, and over 80% publicly owned in the west.^[6]

The efficiency-related boundaries define a range of fuel substitution efficiencies for different biomass-combustion pathways. Different supply chains emit different amounts of carbon per supplied energy unit, and different combustion facilities convert the chemical energy stored in different fuels to heat or electrical energy with different efficiencies. The researcher has to know about this and choose a realistic efficiency range for the different biomass-combustion paths under consideration. The chosen efficiencies are used to calculate so-called "displacement factors" – single numbers that shows how efficient fossil carbon is substituted by biogenic carbon.^[u] If for instance 10 tonnes of carbon are combusted with an efficiency half that of a modern coal plant, only 5 tonnes of coal would actually be counted as displaced (displacement factor 0.5). Schlamadinger & Marland describes how such low efficiency lead to high parity times when bioenergy and coal-based forest protection scenarios are compared, and on the other hand how an efficiency identical to the coal scenario lead to low parity times.^[9] Generally, fuel burned in inefficient (old or small) combustion facilities gets assigned lower displacement factors than fuel burned in efficient (new or large) facilities, since more fuel has to be burned (and therefore more CO₂ released) in order to produce the same amount of energy.^[v]

Likewise, since the production of wood based construction materials demand lower fossil fuel input than the production of fossil based construction materials (e.g. cement or steel), the wood based construction materials get assigned displacement factors when substitution of cement and steel based construction materials is realistic, i.e. when they have the same utility in construction. The more fossil fuel emissions avoided by using utility-equivalent wood construction products, the higher the assigned displacement factors.^[w] Additionally, the carbon stored in wood products during the products' service life, and the fossil carbon that is displaced when the wood products are combusted for energy at the end of their service life, can both be included in the displacement factor calculations. However, so far this is not common practice.^[x] (52% of the harvested forest biomass in the EU is used for materials.)^[10]

Sathre & O'Connor examined 21 individual studies and found displacement factors of between −2.3 and 15 for construction wood products, with the average at 2.1, which means that for each tonne of biogenic carbon produced, on average 2.1 tonnes of fossil carbon is displaced.^[11] For wood based biofuels, the displacement factors varied between roughly 0.5 and 1, "[...] depending largely on the type of fossil fuel replaced and the relative combustion efficiencies."^[12] The authors write that when construction wood products are combusted for energy at the end of their service life, the displacement effect is sometimes added to the calculation, "[...] as the GHG benefits of both material substitution and fuel substitution accrue."^[13] In another meta study on construction wood products, where this additional end-of-life combustion substitution effect was excluded, the authors found somewhat lower displacement factors. The combustion-specific displacement factors were similar but with a wider range (see charts on the right.)^[14]

The displacement factor varies with the carbon intensity of both the biomass fuel and the displaced fossil fuel. If or when bioenergy can achieve negative emissions (e.g. from afforestation, energy grass plantations and/or bioenergy with carbon capture and storage (BECCS),^[y] or if fossil fuel energy sources with higher emissions in the supply chain start to come online (e.g. because of fracking, or increased use of shale gas), the displacement factor will start to rise. On the other hand, if or when new baseload energy sources with lower emissions than fossil fuels start to come online, the displacement factor will start to drop. Whether a displacement factor change is included in the calculation or not, depends on whether or not it is expected to take place within the time period covered by the relevant scenario's temporal system boundaries.^[z]

The economic boundaries define which market effects to include in the calculation, if any. Changed market conditions can lead to small or large changes in carbon emissions and absorptions from supply chains and forests,^[aa] for instance changes in forest area as a response to changes in demand. Miner et al. describe how researchers have begun to examine forest bioenergy in a broader, integrated framework that also addresses market impacts. Based both on empirical data and modeling, these studies have determined that increased demand often leads to investments in forestry that increase forest area and incentivize improvements in forest management. Depending on circumstances, this dynamic can increase forest carbon stocks. Where growth rates are relatively high and the investment response strong, net GHG benefits from increased use of trees for energy can be realized within a decade or two, depending on the fossil fuel being displaced and the timing of the investment response. Where tree growth is slow and the investment response is lacking, many decades may be required to see the net benefits from using roundwood for energy. The investment response has been found to be especially important in places such as the US South, where economic returns to land have been shown to directly affect gains and losses in forest area.^[15] Abt et al. argue that the US South is the world's largest timber producer, and that the forest is privately owned and therefore market driven.^[16] Further, EU's Joint Research Centre argue that macroeconomic events/policy changes can have impacts on forest carbon stock.^[ab] Like with indirect land use changes, economic changes can be difficult to quantify however, so some researchers prefer to leave them out of the calculation.^[ac]

The chosen system boundaries are very important for the calculated results.^[ad] Shorter payback/parity times are calculated when fossil carbon intensity, forest growth rate and biomass conversion efficiency increases, or when the initial forest carbon stock and/or harvest level decreases.^[17] Shorter payback/parity times are also calculated when the researcher choose landscape level over stand level carbon accounting (if carbon accounting starts at the harvest rather than at the planting event.) Conversely, longer payback/parity times are calculated when carbon intensity, growth rate and conversion efficiency decreases, or when the initial carbon stock and/or harvest level increases, or the researcher choose stand level over landscape level carbon accounting.^[ae]

Critics argue that unrealistic system boundary choices are made,^[af] or that narrow system boundaries lead to misleading conclusions.^[ag] Others argue that the wide range of results shows that there is too much leeway available and that the calculations therefore are useless for policy development.^[ah] EU's Join Research Center agrees that different methodologies produce different results,^[ai] but also argue that this is to be expected, since different researchers consciously or unconsciously choose different alternative scenarios/methodologies as a result of their ethical ideals regarding man's optimal relationship with nature. The ethical core of the sustainability debate should be made explicit by researchers, rather than hidden away.^[aj]

According to EU's Joint Research Centre, the use of boreal stemwood harvested exclusively for bioenergy have a positive climate impact only in the long term, while the use of wood residues have a positive climate impact also in the short to medium term.^[ak] See chart on the right for an overview over expected emission reductions from different forest bioenergy pathways, including stemwood, residues and new plantations, compared against energy generation from coal and natural gas in the alternative scenarios. Stems from short-rotation coppices or short-rotation forests also have positive climate effects in the short to medium term (see below.)

Short carbon payback/parity times are produced when the most realistic no-bioenergy scenario is a traditional forestry scenario where "good" wood stems are harvested for lumber production, and residues are burned or left behind in the forest or in landfills. The collection of such residues provides material which "[...] would have released its carbon (via decay or burning) back to the atmosphere anyway (over time spans defined by the biome's decay rate) [...]."^[19] In other words, payback and parity times depend on the decay speed. The decay speed depends on a.) location (because decay speed is "[...] roughly proportional to temperature and rainfall [...]"^[20]), and b.) the thickness of the residues.^[al] Residues decay faster in warm and wet areas, and thin residues decay faster than thick residues. Thin residues in warm and wet temperate forests therefore have the fastest decay, while thick residues in cold and dry boreal forests have the slowest decay. If the residues instead are burned in the no-bioenergy scenario, e.g. outside the factories or at roadside in the forests, emissions are instant. In this case, parity times approach zero.^[am]

Madsen & Bentsen examined emissions from both forest residues and coal, combusted in the same, real Northern European CHP (combined heat and power) plant, and found that the carbon parity time was 1 year.^[an] The low parity time was mainly the result of the use of residues, the generally high conversion efficiencies of CHP plants compared to regular power plants (in this case 85.9%), and longer transport distance for coal.^[ao] The authors note that most bioenergy emission studies use hypothetical rather than real field data, and that 16 times more biomass is combusted in CHP plants than in pure electricity plants in the EU.^[ap] In other words, it is heat-related payback/parity times such as these that are the most relevant for the current situation.

Holmgren studied climate effects from actual forestry practices in a whole country over a 40-year time period (Sweden 1980–2019), and found that at the national landscape level, no carbon debt accrued at any point in time during this period. The actual forestry practice was compared to two alternative forest protection scenarios. The counted emissions caused by the initial harvest in the actual forestry scenario did not lead to a carbon debt because 1.) the initial harvest-related carbon emission was outweighed by carbon absorption caused by growth elsewhere in the forest (a trend that is expected to continue in the future), and 2.) because a national forest protection policy would cause large initial emissions from the national wood-based products and energy infrastructure when it is converted to work with fossil fuels.^[aq] The conversion is described as a "[...] one-off transformation, representing major and required modifications to energy systems, infrastructure, industrial processing, building sector, manufacturing of consumer products and other economic activities towards fossil-based production if a no-harvest scenario were to be implemented."^[21] Of course, if the bioenergy scenario's initial harvest-related emission event is outweighed by 1.) forest growth elsewhere, and 2.) infrastructure conversion emissions (in the forest protection scenario), no carbon debt accrues at all, and the payback and parity times reduce to zero. The author argue that since forest protection most likely will cause fossil carbon to be emitted instead of biogenic carbon, the practical effect of forest protection is simply a transfer of carbon from the underground fossil carbon pool via combustion to the atmospheric carbon pool, and then via photosynthesis to the forest carbon pool. However, when carbon is stored in forests instead of underground fossil reservoirs, it is more unstable, that is, easier to convert to CO₂ because of natural disturbances.^[22] A conservative displacement factor of 0.78 tonnes of fossil carbon displaced per tonne of biogenic carbon produced is used for both harvested wood products (HWP) and energy combined.^[ar] The author criticizes studies that limit carbon accounting to the carbon flows within the forests themselves and leave out fossil displacement effects, and argues that this narrow system boundary essentially works as "[...] a justification for continued fossil emissions elsewhere with no net gain for the global climate."^[23] In Sweden, the biomass that is available for energy is mainly used in heating facilities (7.85 Mtoe used for heating, 0.84 Mtoe for electricity.)^[24]

Zanchi et al. agree that there are climate benefits from the beginning when using easily decomposable forest residues for bioenergy. They also write that "[...] new bioenergy plantations on lands with low initial C [carbon] stocks, such as marginal agricultural land, has the clearest advantages in terms of emission reductions."^[29] The reason is that newly planted areas (which now has a large growing stock of trees or other plants), absorb much more carbon than earlier. Such areas build up a carbon credit instead of a carbon debt, where the credit is used later (at harvest) to acquire "debt free" biomass. In general, "early" carbon accounting like this, which starts at the planting event rather than at the harvest event (cf. Temporal system boundaries above), is seen as uncontroversial for new bioenergy plantations on land areas with very little vegetation. On the other hand, for areas where there already is a large amount of vegetation in place, "late" carbon accounting is often preferred. In this case, carbon accounting starts at harvest, with no build-up of a prior carbon credit. With this type of carbon accounting, the calculated results show that there are short to medium term negative impacts when trees are felled exclusively for bioenergy (so-called "additional fellings"). The situation gets worse if residues are left to rot on the forest floor. There is also a risk for negative impacts if areas with large amounts of biomass such as forests are clear-cut in order to make room for low-productivity forest plantations.^[30]

The assessment of such "additional fellings" from "new" bioenergy plantations after the first rotation is complete, depends on the chosen carbon accounting method. If the "early" carbon accounting continues, there will be a build-up of a carbon credit also after the first rotation, i.e. from the moment in time when the trees have been replanted. If the researcher at that time change to "late" carbon accounting, no carbon credit will be calculated, and at the end of the second rotation (at harvest) a large carbon debt will be created instead, causing payback and parity times to increase dramatically.

EU's Joint Research Centre provides time-dependent emission estimates for electricity production on a large scale from residue-based wood pellets, cereal straw and biogas from slurry, compared against a no-bioenergy scenario with emissions equal to EU's current electricity mix. Conversion efficiencies are 34%, 29% and 36% for wood pellets, straw and biogas, respectively. If not used for electricity production, the forest residues would have been left to decay on the forest floor, the straw residues would also have been left behind in the fields, and the raw manure would have been used as organic fertilizer. The results show that if these biomass types instead were used to produce electricity, a global warming mitigation effect would start after approximately 50, 10 and 5 years of use, for wood, straw and biogas respectively. The main cause for the long parity time for wood pellets is the comparison with electricity from EU's electricity mix (which includes electricity from solar, wind and fossil fuels with lower emissions than coal). Also, the forest residue category includes stumps.^[aw]

The JRC also found parity^[ax] times ranging from 0 to 35 years for harvest residues (including branches, thinnings and stumps), when compared to some other alternative scenarios. In Finland, parity times for stumps were 22 years compared against oil, and 35 years compared against natural gas, with stand level carbon accounting. In Canada, parity time increased from 16 to 74 years when the harvested biomass was used to produce ethanol instead of wood pellets, and compared against a gasoline-based alternative scenario instead of a coal-based alternative scenario.^[ay] Ethanol production from whole trees removed from old-growth forests in Oregon, USA, (categorized as residues because the trees were felled to prevent wildfire) increased parity time dramatically, with the worst-case scenario at 459 years. The authors used stand level carbon accounting starting with the harvest event, assumed an additional, controlled burning every 25 years, and compared this to a scenario with no wildfire-preventive fellings and a severe wildfire every 230 years.^[31] The trees in question were huge western hemlock and coast Douglas fir trees which both take hundreds of years to mature and can withstand wildfires due to very thick stems. Since energy-intensive ethanol production caused a low displacement factor of only 0.39, a long parity time was calculated.^[32] Generally, the JRC's reported parity times were influenced by displacement factor, alternative scenario, residue size and climate type. See chart above.

If an existing natural forest is clear-cut in order to make room for forest plantations, the implied carbon change create a significant carbon debt roughly equal to the amount of carbon residing in the felled trees (fossil based forestry operations create an additional, small debt.) But for new plantations on "empty" land like agricultural or marginal land, with no standing vegetation, no carbon is removed. In this case, a carbon credit is instead soon built up as the trees mature. When those trees later are felled, the amount of carbon that resides in the trees is subtracted from the built up carbon credit (not the carbon amount in the standing trees), so in this case no carbon debt is created. With no carbon debt created at harvest, carbon payback/parity times will be zero or very low, for residues and stemwood alike.^[az]

Jonker et al. calculated both carbon payback and carbon parity times for stemwood with rotation times of 20–25 years harvested from southeastern forests in the US, using both stand level, increasing stand level, and landscape level carbon accounting. With stand-level carbon accounting, the authors found carbon payback times of 5, 7 and 11 years in the high, medium and low yield scenario, respectively. With increasing stand level accounting, the payback times were 12, 13 and 18 years in the high, medium and low yield scenario, respectively. With landscape level accounting, the payback time was below 1 year for all yield scenarios.^[3] The authors also calculated parity times for a scenario where wood pellets from stems only (no residue collection) were used for co-firing in an average, coal-based electricity plant. The conversion efficiency was 41%, which together with an efficient supply chain leads to a relatively high displacement factor of 0.92. The alternative scenario was a no-bioenergy scenario where the stemwood was instead used for lumber production, so no co-firing at all in this case (electricity from coal exclusively.) When using the increasing stand level accounting principle, the authors calculated parity times of 17, 22 and 39 years for the high, medium and low yield scenario, respectively. When using the landscape level accounting principle, the authors calculated parity times of 12, 27 and 46 years for the high, medium and low yield scenario, respectively. A different alternative scenario was a forest protection scenario where no biomass was extracted from the forest at all; not for lumber, and not for bioenergy. The forest was simply left to itself and therefore regrew rather slowly. Landscape level parity times for this scenario was 3, 3, and 30 years for the high, medium and low yield scenario, respectively (stand level or increasing stand level parity times were not provided.)^[33]

Zanchi et al. found that parity times can reach 175 years with a coal-based alternative scenario and 300 years with a natural gas-based alternative scenario if spruce stems in the Austrian Alps are harvested exclusively for bioenergy. The main reason is the long rotation time for these trees (90 years). Generally, trees take 70–120 years to mature in boreal forests.^[38] Critics reply that stems that meet quality requirements are used to produce high-value products such as sawn wood and engineered wood products such as cross laminated timber, rather than low-value products such as wood pellets.^[be] In a different scenario where forests of this type is clear-cut and used 50/50 for bioenergy and solid wood products, and then subsequently replaced with short rotation forest, parity times varies between 17 and 114 years for the coal alternative scenario, with the shortest parity time achieved by the forest with the shortest rotation time and highest yield (10 years rotation time with a yield of 16 tonnes per hectare per year.) Parity times increased to between 20 and 145 years when compared to an oil-based electricity alternative case, and between 25 and 197 years when compared to a natural gas-based electricity alternative case. For an afforestation vs. fossil fuel mix scenario, a parity time of 0 years was reported.

The authors note that these scenarios are "illustrative examples" and that "results are strongly influenced by the assumptions made." The authors assumed that residues were left un-collected on the forest floor, where they decay and therefore produce emissions. If these residues instead are collected and used for bioenergy, the parity times decrease by 100 years. The extra emissions produced by the longer supply routes for fossil fuels compared to wood fuel were not included in the calculation.^[bf] Extra emissions from pests, windthrows and forest fires (normally expected to increase when unmanaged forests age), were also not included in the calculation. Market effects were not included. On the other hand, landscape level carbon accounting was used, and the assumed conversion efficiency for bioenergy and coal were the same.^[39]

Like other scientists, the JRC staff note the high variability in carbon accounting results, and attribute this to different methodologies.^[bg] In the studies examined, the JRC found carbon parity times of 0 to 400 years (see chart on the right) for stemwood harvested exclusively for bioenergy, depending on different characteristics and assumptions for both the forest/bioenergy system and the alternative fossil system, with the emission intensity of the displaced fossil fuels seen as the most important factor, followed by conversion efficiency and biomass growth rate/rotation time. Other factors relevant for the carbon parity time are the initial carbon stock and the existing harvest level; both higher initial carbon stock and higher harvest level means longer parity times.^[40] Liquid biofuels have high parity times because about half of the energy content of the biomass is lost in the processing.^[bh]

This section may need to be rewritten to comply with Wikipedia's quality standards. You can help. The talk page may contain suggestions. (January 2023)

EU's Joint Research Centre has examined a number of bioenergy emission estimates found in literature, and calculated greenhouse gas savings percentages for bioenergy pathways in heat production, transportation fuel production and electricity production, based on those studies (see charts on the right). The calculations are based on the attributional LCA accounting principle. It includes all supply chain emissions, from raw material extraction, through energy and material production and manufacturing, to end-of-life treatment and final disposal. It also includes emissions related to the production of the fossil fuels used in the supply chain. It excludes emission/absorption effects that takes place outside its system boundaries, for instance market related, biogeophysical (e.g. albedo), and time-dependent effects. Because market related calculations are excluded, the results are only seen as valid for small-scale energy production.^[43] Also, the bioenergy pathways have typical small-scale conversion efficiencies. Solid biofuels for electricity production have 25% efficiency in most cases, and 21–34% in a few cases. Biogas for electricity production have 32–38%. Heat pathways have 76–85%. The forest residue category include logs and stumps, which increases carbon intensity especially in forests with slow decay rates.^[44]

The charts have vertical bars that represent the emission range found for each bioenergy pathway (since emissions for the same pathway vary from study to study.) The higher end of the range represents emission levels found in studies that assume for instance long transport distances, low conversion efficiencies and no fossil fuel displacement effect. The lower end of the range represents emission levels found in studies that assume optimized logistics, higher conversion efficiencies, use of renewable energy to supply process-heat and process-electricity, and include displacement effects from the substitution of fossil fuels.^[bj] The bars can be compared with emission levels associated with multiple alternative energy systems available in the EU. The dotted, coloured areas represent emission savings percentages for the pathways when compared to fossil fuel alternatives.^[42] The authors conclude that "[m]ost bio-based commodities release less GHG than fossil products along their supply chain; but the magnitude of GHG emissions vary greatly with logistics, type of feedstocks, land and ecosystem management, resource efficiency, and technology."^[45]

Because of the varied climate mitigation potential for different biofuel pathways, governments and organizations set up different certification schemes to ensure that biomass use is sustainable, for instance the RED (Renewable Energy Directive) in the EU and the ISO standard 13065 by the International Organization for Standardization.^[46] In the US, the RFS (Renewables Fuel Standard) limit the use of traditional biofuels and defines the minimum life-cycle GHG emissions that are acceptable. Biofuels are considered traditional if they achieve up to 20% GHG emission reduction compared to the petrochemical equivalent, advanced if they save at least 50%, and cellulosic if the save more than 60%.^[bk]

Consistent with the charts, the EU's Renewable Energy Directive (RED) states that the typical greenhouse gas emissions savings when replacing fossil fuels with wood pellets from forest residues for heat production varies between 69% and 77%, depending on transport distance: When the distance is between 0 and 2500 km, emission savings is 77%. Emission savings drop to 75% when the distance is between 2500 and 10 000 km, and to 69% when the distance is above 10 000 km. When stemwood is used, emission savings varies between 70% and 77%, depending on transport distance. When wood industry residues are used, savings varies between 79% and 87%.^[bl]

Based on a similar methodology, Hanssen et al. found that greenhouse gas emissions savings from electricity production based on wood pellets produced in the US southeast and shipped to the EU, varies between 65% and 75%, compared to EU's fossil fuel mix.^[bm] They estimate that average net GHG emission from wood pellets imported from the US and burnt for electricity in the EU amounts to approximately 0.2 kg CO₂ equivalents per kWh, while average emissions from the mix of fossil fuels that is currently burnt for electricity in the EU amounts to 0.67 kg CO₂-eq per kWh (see chart on the right). Ocean transport emissions amounts to 7% of the displaced fossil fuel mix emissions per produced kWh.^[bn]

Likewise, IEA Bioenergy estimates that in a scenario where Canadian wood pellets totally replace coal in a European coal plant, the ocean transport related emissions (for the distance Vancouver – Rotterdam) amounts to approximately 2% of the plant's total coal-related emissions.^[47] The lower percentage here is caused by the alternative scenario being a particular coal plant, not EU's fossil fuel mix. Cowie et al. argue that calculations from actual supply chains show low emissions from intercontinental biomass transport, for instance the optimized wood pellet supply chain from the southeastern USA to Europe.^[bo] Lamers & Junginger argue that future EU import of wood pellets "[...] will likely continue to be dominated by North America, especially from the South-East USA [...]."^[48] In 2015, 77% of the imported pellets came from the USA.^[bp]

While regular forest stands have rotation times spanning decades, short rotation forestry (SRF)^[bq] stands have a rotation time of 8–20 years, and short rotation coppicing (SRC)^[br] stands 2–4 years.^[49] 12% of the EU forests is coppice forests.^[bs] Perennial grasses have a rotation time of one year in temperate areas, and 4–12 months in tropical areas.^[50] Food crops like wheat and maize also have rotation times of one year.

Because short rotation energy crops only have managed to grow/accumulate carbon for a short amount of time before they are harvested, it is relatively easy to pay back the harvest-related carbon debt, provided that there is no additional large carbon debt from land use change to deal with (for instance created by clear-cutting a natural forest in order to use this land area for energy crops), and no better climate-related use for the areas in question. Schlamadinger & Marland write that "[...] short-rotation energy crops will provide much earlier and larger C [carbon] mitigation benefits if implemented on previously unforested land than if an initial forest is harvested to provide space for the plantation."^[51] EU's Joint Research Centre state: "In case that there is no raw material displacement from other sectors such as food, feed, fibers or changes in land carbon stocks due to direct or indirect land use change, the assumption of carbon neutrality can still be considered valid for annual crops, agriresidues, short-rotation coppices and energy grasses with short rotation cycles. This can also be valid for analysis with time horizons much longer than the feedstock growth cycles."^[52] Other researchers argue that the small carbon debts associated with energy crop harvests means short carbon payback and parity times, often less than a year.^[bt] IRENA argues that short-rotation energy crops and agricultural residues are carbon neutral since they are harvested annually.^[bu] IEA writes in its special report on how to reach net zero emissions in 2050 that the "[...] energy‐sector transformation in the NZE [Net Zero Emissions scenario) would reduce CO₂ emissions from AFLOU [Agriculture, Forestry and Other Land Use] in 2050 by around 150 Mt CO₂ given the switch away from conventional crops and the increase in short rotation advanced‐bioenergy crop production on marginal lands and pasture land."^[53]

Typically, perennial crops sequester more carbon than annual crops because the root buildup is allowed to continue undisturbed over many years. Also, perennial crops avoid the yearly tillage procedures (plowing, digging) associated with growing annual crops. Tilling helps the soil microbe populations to decompose the available carbon, producing CO₂.^[by][bz] Soil organic carbon has been observed to be greater below switchgrass crops than under cultivated cropland, especially at depths below 30 cm (12 in).^[54] A meta-study of 138 individual studies, done by Harris et al., revealed that the perennial grasses miscanthus and switchgrass planted on arable land on average store five times more carbon in the ground than short rotation coppice or short rotation forestry plantations (poplar and willow).^[ca] McCalmont et al. compared a number of individual European reports on Miscanthus × giganteus carbon sequestration, and found accumulation rates ranging from 0.42 to 3.8 tonnes per hectare per year,^[cb] with a mean accumulation rate of 1.84 tonne,^[cc] or 25% of total harvested carbon per year.^[cd]

This above-ground circulation has the potential to be carbon neutral, but of course the human involvement in operating and guiding it means additional energy input, often coming from fossil sources. If the fossil energy spent on the operation is high compared to the amount of energy produced, the total CO₂ footprint can approach, match or even exceed the CO₂ footprint originating from burning fossil fuels exclusively, as has been shown to be the case for several first-generation biofuel projects.^[cf][cg][ch] Transport fuels might be worse than solid fuels in this regard.^[ci]

Successful storage is dependent on planting sites, as the best soils are those that are currently low in carbon.^[ck] For the UK, successful storage is expected for arable land over most of England and Wales, with unsuccessful storage expected in parts of Scotland, due to already carbon rich soils (existing woodland). Also, for Scotland, the relatively lower yields in this colder climate makes carbon negativity harder to achieve. Soils already rich in carbon include peatland and mature forest. The most successful carbon storage in the UK takes place below improved grassland.^[cl] However, since the carbon content of grasslands vary considerably, so does the success rate of land use changes from grasslands to perennial.^[cm] Even though the net carbon storage below perennial energy crops like miscanthus and switchtgrass greatly exceeds the net carbon storage below regular grassland, forest and arable crops, the carbon input is simply too low to compensate for the loss of existing soil carbon during the early establishment phase.^[56] Over time however, soil carbon may increase, also for grassland.^[57]

Researchers argue that after some initial discussion, there is now (2018) consensus in the scientific community that "[...] the GHG [greenhouse gas] balance of perennial bioenergy crop cultivation will often be favourable [...]", also when considering the implicit direct and indirect land use changes.^[cn]