Relative CO2 emissions from power stations fired by wood or coal –

Comments on the model by Peter Barnes

 

July 2001

 

 

 

1.     Summary

An unpublished model constructed by Peter Barnes compares gross and net CO2 emissions from wood- and coal-fired power stations (PS), based on a recent proposal by the NSW Forest Products Association (FPA).  Based on outputs from the model Barnes claimed that ‘while on native-forest fuel, a wood PS causes about 5 times the CO2 output of a coal PS’, and that ‘for power supply, native forest is not CO2 neutral, it is far worse than coal’.

 

The model has several critical flaws, assumptions, and errors of logic that invalidate these conclusions.

 

The main problems with the model are that it:

 

§         Assumes native old growth is harvested and replaced by plantation on a 25-yr rotation.  The stems only are removed to fire the PS, residues being left on site to decay.  Under existing NSW and Commonwealth legislation this is an almost impossible scenario. It is also unlikely because of price competition by high value veneer, sawlogs, pulpwood and composite board markets.

§         The assumed rate of growth of the plantation is excessively low (6.5 m3 ha-1 yr-1), less than half that for typical managed plantations.

§         CO2 output (Sheet 3) is calculated using gross emissions only, including those from decaying residues left on site.  That is, no uptake of CO2 by the regrowing forest is counted. In any case, counting gross emissions is inappropriate because even undisturbed forests at equilibrium (0 net emissions) emit a gross amount of CO2.

§         The expansion factor (ratio of total biomass to stem biomass) used for native forest is much too high, by nearly a factor 2.

§         Incorrect values have been used for parameters to calculate the energy yield from wood, namely low heating values and low energy conversions compared with coal.

 

Using correct parameter values in the model, and an appropriate and legal, renewable forest system in NSW, it is shown that burning coal releases about 8 times more CO2 into the atmosphere than a forest managed for sawn timber and where wastes and residues are used for power generation.  Recent overseas studies have shown that the net emission of CO2 from a coal-fired PS is about 30 times that from a wood-fired PS (DTI, 1999).

 

Three basic principles need to be kept in mind when comparing net emissions of CO2 derived from forests and coal to generate electricity:

 

§         Fossil fuels are not renewable so that when coal or gas is burnt there is a gross emission of CO2 that can never be removed from the atmosphere by the fossil fuel.

§         When a forest is harvested or burnt by wildfire and the wood used for any purpose, there is initially a net emission of CO2 that progressively declines as the stand ages and uptake by the regrowing vegetation dominates.  At the end of the rotation the net emission of CO2 is close to zero (Polglase et al. 1994).

§         When some of the wood from a harvested or burnt forest is used to generate electricity, substituting for coal, an additional benefit from harnessing this renewable energy source is gained and less CO2 remains in the atmosphere than would otherwise be the case because the coal is left in the ground.

 

This basic framework, apart from being self-evident, has been quantified by credible life cycle assessments that have compared CO2 emissions from wood- and coal-fired power stations.  The contrary model of Barnes should be submitted to a scientific journal for peer review before being cited as evidence for the role of forest-based bioenergy in greenhouse gas balances.

 

 

2.     Background

The analysis of Barnes attempts to compare gross and net CO2 emissions from wood- and coal-fired power stations (PS), based on a recent proposal by the NSW Forest Products Association.  The model claimed that:

 

§         ‘While on native-forest fuel, the wood PS causes about 5 times the CO2 output of a coal PS’,

§         ‘By the end of its 80-yr life cycle, wood power leaves… nearly 3 times the CO2 emissions of a coal power station for the same amount of electricity’,

§         Realistically, for power supply, native forest is not CO2 neutral, it is far worse than coal.

 

An Excel spreadsheet was developed by Barnes and disseminated to various parties.  The NSW Nature Conservation Council issued a press release on 24 May before the model was published or underwent critical assessment.  The press release included the comments:

 

§         ‘The community is vehemently opposed to the exploiting of native forest for electricity generation, and this report justifies the level of hostility from the public towards these proposal’,

§         ‘Defenders of the plan have tried to stretch scientific arguments to justify unsustainable practices’, and that

§         ‘The facts are that wood is an inefficient source of power and burning it for this purpose will release more CO2 into the atmosphere than existing coal fired stations’.

 

These statements were made without qualification.

 

The aim here is to objectively and independently review the scientific framework and assumptions of the Barnes model.  The version of the model reviewed is ‘Wood Pwr in XL-97 Rev.H.xls’ although we note that subsequent versions of the model have been constructed, with much the same output as in the ‘H’ version.

 

3.     What the model does

The model is based on the proposal by the FPA that 3 power stations be built in NSW, fired by biomass, each with a generating capacity of 30 MW.  It is important to recognise the assumptions underlying the Barnes model and key parameters that were used:

 

Assumptions

 

§         The model ignores the Commonwealth and State legislative framework required by the Renewable Energy (Electricity) Act (2000) and its ESD provisions, and the Native Vegetation Conservation Act (1997).

§         A wood-fired PS is built, and the emissions associated with that construction are debited against the wood PS in a partial life-cycle assessment.  There is no such debiting for the coal station because the coal-fired power station is assumed to already exist.

§         In the first 25 years an area of native forest, which is assumed to be more or less in equilibrium, is logged.  Stems only are taken to the wood-fired power station and burnt.  Residues (branches, leaves, and roots) are left behind to decompose.

§         Each area of native forest harvested is then replaced with eucalypt plantation on a 25-yr rotation.  After the first 25 years this plantation then begins to be used as the biomass resource.

§         The rate of plantation growth is slow, having a mean annual volume increment of 6.5 m3 ha-1 yr-1 over 25 years.

§         When plantations come on line, they are harvested in preference to native forest, enough native forest then being harvested to make up the difference.  This continues until the whole native forest area has been converted to plantation (effectively delineating the project boundary).

§         Stems only in the plantation are taken to the wood-fired power station and burnt. Residues (branches, leaves, and roots) are left behind to decompose.

§         The project has a life-span of 80 years, which means no more burning of wood for power generation.  However, after 80 years, the plantations continue to be harvested (keeping their carbon density low) and net emissions are then debited against the wood-fired power station.

§         The first commitment period under the Kyoto Protocol is used as a reference for net emissions from the wood- and coal-fired PSs.

 

Parameters

§         The calorific values  (MJ kg-1) are 10 for wood (35% moisture content) and 20 for black coal.

§         The efficiencies for conversion (%) are 15 for wood, 27 for brown coal, and 36 for black coal.

§         An expansion factor of 2.5 is used to convert stem biomass to total biomass for native forest, and a factor of 1.6 is used for plantations.

 

These validity of these assumptions and parameter values is considered below.

 

 

4.     What the model should do

 

CO2 Output (Sheet 3)

The way in which this is calculated is invalid, and is the basis for the comment that the “wood PS causes about 5 times the CO2 output of a coal PS’,

 

It compares the gross CO2-output associated with coal with that generated from harvesting (diesel use) and burning wood (PS emissions).  For the wood PS it includes emissions from residues left decaying on the forest floor.  This in particular is inappropriate because:

 

i.                     There is no allowance for any replacement by the forest of the carbon emitted, and

ii.                   These residues could be utilised by the wood PS.

 

In undisturbed forest at equilibrium, CO2 is constantly emitted by the decay of litterfall (leaves, branches), sloughed roots, and from fallen whole stems (mortality).  Counting gross CO2 emissions from litter decay in the model, without matching uptake, is like counting gross emissions from the undisturbed mature forest.  These typically might be 18 t CO2  ha-1 yr-1 but are matched by an equal amount in the opposite direction so that net emissions are zero.

 

It also compares gross CO2 output over the first 25 years of the project.  Why was this period chosen when gross emissions from the coal- and wood-fired stations are still rising – that is they are on lines of different slopes?  It would make more sense to consider net emissions at the end of the project, and even better sense to consider the time frame of a forest cycle.

 

No allowance had been made for the embedded CO2 emissions associated with building successive power stations if the assessment period is greater than the life of a one-off power station.

 

The only sensible way to compare gross CO2 output between coal and wood is by using conversion efficiency and specific CO2 yields of the respective fuels.

 

In order to generate 1GWh of electricity, the correct formula would be:

 

Gross CO2 output/unit of electrical energy produced = (CO2/energy) / conversion efficiency

 

Thus the CO2 out of the power station stack in order to generate 1 GWh of electricity for each of the fuels is:

 

 

CO2/energy

(kt CO2 PJ-1)*

Conversion efficiency (%)

CO2 output

(kt CO2 GWh-1)

 

 

 

 

Wood

94

20

1.69

Brown coal

93

27

1.24

Black coal

92

36

0.92

*from NSW Department of Energy (now NSW MoU, 2000).

 

Note that the values for specific CO2 yield and conversion efficiency have been revised from the original Barnes model (see discussion below).

 

Thus it is generally true that burning wood releases more CO2 per unit of electricity generated than burning coal.  But because coal is not a renewable resource, in terms of greenhouse gas balances and what the atmosphere sees, it is much more important to include complete life cycle assessment to calculate net CO2 emissions over the life cycle of a forest.  Sheets 4 and 5 of the Barnes model attempt to do this, but do so incorrectly.

 

The amount of CO2 that remains in the atmospheric is directly related to the conversion of fossil carbon into atmospheric CO2. Carbon is sustainably managed forest is an integral part of the dynamic non-anthropogenic carbon exchanges with the atmosphere. Harnessing the energy flows from this renewable resource for power generation is therefore carbon neutral.

 

Conversion of old growth to plantation

The Barnes analysis assumes permanent conversion of forest from high carbon density to low carbon density.  This is a form of land-use change and, in part, accounts for the disparity between CO2 emissions from the wood and coal PSs.  This assumption is incorrect on several fronts:

§         This was not part of the FPA proposal.  Again, they proposed only to use residues and mill waste from sustainably managed forest (forest that returns to equilibrium).

§         Under NSW legislation (Native Vegetation and Conservation Act ) clearing of native forest is strictly controlled and the assumptions in the model are unlikely at any significant scale.  The Regional Forest Agreement has also specified areas of public native forest that are to be sustainably managed.

§         Over the time scale of centuries, forest will always have the potential to return to equilibrium, in contrast to fossil fuels sources (and see discussion below under ‘Carbon balances in forests and implications for the Barnes model’).  Over the past 160 years the concentration of CO2 in the atmosphere has risen from about 290 parts per million (ppm) to about 370 ppm.  Thus sudden rise has been due mostly to the burning of fossil fuels, but also to the permanent conversion of forest to cultivated land.  Emissions from forest fires and sustainably managed forest have contributed little, the CO2 balance returning to zero.

§         The carbon stock of the entire forest estate needs to be considered, not just patches of forest.

 

Rate of plantation growth

This is very low, being only 6.5 m3 ha-1 yr-1 over 25 years.  A more realistic value would be at least 15 m3 ha-1 yr-1 for a reasonably managed forest.  This translates into a ‘log density at harvest’ (stem mass) of 300 t ha-1.

To be consistent, for the native forest a log density at harvest of 350 t ha-1 yr-1 could be assumed.

 

Using only stems, but not residues, in the  wood PS

This assumption is incorrect (see above on ‘Conversion of old growth to plantation’).  Under current Renewable Energy legislation only residues can be used.  The FPA stated that only mill waste and in some cases residues on the forest floor, left after harvesting, would be utilised in a wood PS.  Therefore the analysis cannot be based on use of stems. Also, the use of stems is unlikely because of price competition by high value veneer, sawlogs, pulpwood and composite board markets.

 

Expansion factor

The model is very sensitive to the expansion factor.  The value of 2.5 used for native forest is much too high.

The expansion factor is defined by Barnes as:

(carbon in total living biomass) / (carbon in stems).

Barnes used a CSIRO report (Snowdon et al., 2000) for deriving the value, stating that a ‘value of 2.5 seems conservative as CSIRO reports up to 5.0’.

 

The CSIRO report defined the expansion factor as:

 

(carbon in total above-ground biomass cut) / (carbon used commercially).

 

The value of 5 cited was for plantations only where there is low recovery of stems for pulp wood.  The expansion factor generally decreases from high values of 3 to 4 in plantations less than 5 years old to a value not much more than 1 at maturity (Snowdon et al., 2000).  For mature native forest there are few good data for expansion factors.  However, as a tree matures the long-lived stem tends to dominate and becomes an increasingly high proportion of the biomass.

 

The Snowdon et al. (2000) report notes that the average value for expansion factor (as they defined it) was about 1.4 for native forest but they noted that this was possibly based on incomplete measurement of the stem.  Note also that this excludes the contribution by roots.  The root:shoot ratio is about 0.2 to 0.25 in forest, but again can be expected to be lower than this in mature forest.

Using the above information and modelling studies (Polglase et al., 1994) an expansion factor of 1.3 to 1.5 is appropriate for mature forest, including roots.

 

Use of uptake by old growth in calculations

This is inappropriate, the rate of uptake by old-growth forest (Un) should be set to 0.  It is not reasonable to count uptake by an area of unharvested forest (or unmanaged for that matter) in any calculation.  This effectively sets an inappropriate boundary for the project.  See later comment under ‘Carbon balances in forests and implications for the Barnes model’.

 

Counting emissions from construction of the wood PS

There was no such adjustment for the coal PS as it was assumed that the coal-fired power station already exists.  In terms of what the atmosphere sees over time, there was a cost (emission) associated with building the coal PS in the first instance, and that CO2 is effectively still in the atmosphere.  In order to be consistent, and compare like with like, the emission cost of building the coal PS, at least on a pro-rata basis (ie 30 MW worth of the coal-fired station), should also be factored in, or emissions associated with building the wood PS not counted. This has now been corrected in the latest version of the model available.

 

Net (not gross) calorific value (ie LHV)

The model is very sensitive to this parameter. For coal in the Barnes model, a gross calorific value (HHV) has been used. For consistency therefore, HHV will be used throughout.  A value of 10 MJ kg-1 is given for the net calorific value (LHV, 35% moisture content) in the Barnes model, but this should be reset to 13 MJ kg-1 (gross calorific value, ie HHV).

 

Electricity generation efficiency

The model is again very sensitive to this parameter.  A value of 15% is given for wood, but is at the low end of the range.  Efficiencies of at least 20% would be appropriate for a 30 MW unit, and up to 27%. There is no one “correct” value – plant efficiency depends on the capital that the developers are prepared to spend, condensor conditions, etc. It is important that if HHV calorific value is used for the fuel, a comparable HHV boiler efficiency is assumed. Alternatively, if LHV calorific value is used, a comparable LHV boiler efficiency should be used. Either method will give the same end result if consistency is maintained. At this capacity, steam conditions are less than is typical for large coal-fired power stations, and reheat is not an economic option, so steam cycle efficiency suffers as a result. For example, temperatures may be of the order of 500oC instead of 540 oC. A condensing turbine is assumed, although the temperature of the cooling medium this will also have a significant effect on efficiency.   Higher efficiencies are possible, but at a higher capital cost, and lower efficiencies at a lower capital cost. In this paper, 20% (on an HHV basis) is assumed, erring on the conservative side. In practice, 23% (HHV) might be a more typical figure, which is roughly equivalent to 27% on an LHV basis at 35% moisture.

 

 

CO2/ energy

This is a derived value in the model. In fact wood has a fairly consistent carbon content on a dry basis and %component depends on free moisture content (ash content is low and fairly constant). CO2 emissions are then simply a stoichiometric calculation from conversion of C to CO2, assuming 100% combustion efficiency (in practice probably about 99%, depending on C burn-out).  It is therefore better to use data from the NSW Emissions Workbook which specifies CO2 emissions from wood and wood waste of 94 kt of CO2 released per PJ of wood fuel combusted. This compares to a derived value of 122 kt PJ-1 in the spreadsheet.

 

It is interesting to note that, in the model, when 13 MJ kg-1 is used for calorific value and 20% for efficiency, the derived yield is 92  kt CO2 PJ-1, in line with the expected value of 94  kt CO2 PJ-1.

 

Kyoto commitment period as a reference point

It is not relevant in the Barnes model to compare net emissions from wood- and coal-fired PSs during the first commitment period of the Kyoto protocol.  This is because emissions from bioenergy projects are not counted under the terms of the Kyoto protocol, in recognition that forests are a renewable resource and that bioenergy projects have net emissions close to 0.

 

5.     Revised parameter values

Given the complexity of the workbook and the equations contained within, it is not possible to ensure that changing parameters retains internal consistency without having to change other equations or parameters.  Nonetheless, changing parameters as suggested below is enlightening.  The parameters apply to the Life Cycle Assessment (LCA), they do not apply the calculation of gross CO2 output which, as mentioned above, has incorrect assumptions embedded within it.  Furthermore, at this stage it assumed that the general forest system of replacing mature forest with plantation is retained.  A more realistic forest regime and its outputs is given at the end of this paper. 

 

Using the revised parameter and input values below, net emission after 100 years from the wood PS is nearly one-fifth (20%) of that from coal PS (see Figure below).  In the original version of the model it was calculated that burning wood emitted 2.5 times more CO2 than coal after 100 years.  Therefore this is a turnaround by a factor of about 13 (2.5/ 0.2), keeping the same basic forest management system assumed by Barnes which, in any case is unrealistic.

 

Parameter

Old value

New value

Cell reference

Net calorific value

10

11

Sheet 4. E6

Or Gross calorific value (MJ/kg)

 

13

Sheet 4. E6

Conversion efficiency (%)

15

20

Sheet 4. E8

Expansion factor for native forest

2.5

1.5

Sheet 4. E22

Native forest log density at harvest (t ha-1)

250

350

Sheet 5. C9

Plantation log density at harvest (t ha-1)

126

300

Sheet 5. H9

Native forest uptake (t CO2 ha-1 yr-1)

1

0

Sheet 4. E26

Post Life-cycle   Plant'n Harvest (%)

90

0

Sheet 4. E17

CO2 emission from construction of wood PS (kt CO2)

2879

0

Sheet 5. I33

 

 

Net emissions of CO2 with revised parameter values.  This simulation retained the assumptions that old growth forest is replaced by plantation, and that stems are burnt in the PS but residues left in the forest.

 

 

6.     Carbon balances in forests and implications for the Barnes model

The Barnes analysis assumes an area of native forest in equilibrium, defines this as the project boundary, and includes only uptake from this area in its calculation of CO2 balances.  This basic assumption is unrealistic because it ignores forest dynamics and greenhouse gas balances across a forest estate.

 

Forests at the landscape level (several hundreds or thousands of square km) are in various stages of growth.  Some patches of forest will be mature (0 net CO2 emission), some will be actively growing having been disturbed (+ve CO2 uptake), and some (a small proportion in Australia) will be old and senescing (-ve CO2 uptake).  Note that the state of these forests will be equally due to management (eg harvesting) as it is to natural disturbances (eg fire) or their absence.  The National Greenhouse Gas Inventory (NGGI) considers Australia’s forest to be a net sink for CO2.

 

It is more realistic to consider project boundaries across the forest estate where disturbance and uptake are approximately equal, and where harvesting causes no net emission of CO2, whether or not the wood is used for electricity generation.  In the Barnes analysis any disturbance to a forest where wood is used for power generation is counted, incorrectly, as a net positive emission.  This concept is explored explicitly in the following section.

 

 

7.     A realistic forest model for CO2 emissions from harvested forest, with use of residues and waste for bioenergy

 

One scenario for a forest model would have the following assumptions (and simplifications).

 

§         The project life is 80 years

§         100 year rotation in regrowth forest

§         Soil carbon does not change

§         The forest is harvested for saw logs, all logs being removed from the forest

§         The recovery of a sawn product is 35% from a log, the rest going to off-cuts and saw dust

§         Wood products have a mean turnover time of 100 years

§         All residues left lying on the forest floor are removed and used for energy generation

§         The site is resown (ie not coppiced), an as such only roots are left on site to decompose

§         Off-cuts and saw dust are used for energy generation

§         Root residues (mainly coarse woody roots) have a mean turnover time of 10 years

§         After harvest the forest is regenerated and returns to its previous carbon density

§         For a forest of reasonable growth we assume that the total amount of dry matter at equilibrium (100 years) is 360 t ha-1

§         Wood is 35% moisture on a wet basis as per the Barnes analysis, although 50% would be more appropriate.

§         The total amount of wood required is 280 kt year-1

§         The expansion factor for the mature native forest is 1.3.

 

There a number of important points in this analysis:

 

§         It accounts for all CO2 emissions from a harvested forest, whether wood is used for power generation or not.

§         A similar pattern of emission would be obtained whether wood was used for pulp, power, or sawn timber.  This is because the recovery of sawn timber from wood is low and the sawdust and off-cuts are usually burnt anyway, or landfilled.

§         When forest products are not used for bioenergy (ie. exclusively for pulp or sawn timber, or disposed of) there is a net build-up of about 16 Mt of CO2 because coal is used as the only source for electricity generation.  When residues and mill waste are used for electricity generation, the net emission from the forest/ wood PS system is about 1 Mt of CO2 after about 150 years, saving about 15 Mt of CO2 from accumulating in the atmosphere.

§         It therefore follows that, to limit the net emission of greenhouse gases into the atmosphere, it is better to use as much wood as possible in a PS (either alone, with co-generation, or in co-firing).

§         The high rate of emission from the forest is also partly because of the assumed diesel expansion factor (0.2).

 

 

 

An example of net CO2 emissions (build-up) from coal and wood-fired power stations. In this scenario almost all the available wood is used at the power station except for 35% of the stem which is used for sawn timber.  The pattern of emission would be typical for any forest whether burnt by wildfire or harvested for pulp, sawn timber or bioenergy.  The scenario is also for any given patch of forest.  In reality, forests across the landscape are either absorbing or releasing CO2 so that the net emission across large areas and at the time scale of forest cycles is about 0 (see ‘Carbon balances in forests and implications for the Barnes model’).

 

 

8.     Economics

·        The Barnes model calculates that the unit cost of power would be about 22 c kWh-1 from the wood-fired PS.  It is not our intention to review in detail the comparative comments of the Barnes economic analysis - the economic viability of any enterprise will be determined by more thorough analysis by the project developer.  However, in this section we present some indicative industry data for wood-fired PSs.

·        The Barnes document creates difficulties in taking data from a number of different sources. Does the 15% efficiency taken from the CSIRO/SEDA study correlate with the plant design of FPA?  The capital cost of an installed 30 MW plant of $40 M may be low, $1500 kW-1 would be more typical, including external plant such as switchyards.  It is not known if the FPA have included some fuel processing in the price of fuel rather than as capital at the power station for example.  This would account for the $40 t-1 fuel price in the Barnes model, which is high when about a third of the resource is coming from sawmill and chipmill residues. The point should be made again that a low capital cost plant with low efficiency could give the same economic outcome as a high capital cost plant with higher efficiency. This is an optimisation that needs to be undertaken by the project developers who have a better idea of fuel and operation and maintenance costs, and the trade-offs between capital and efficiency.

·        $80 t-1 is too high for the cost of black coal. In NSW, the marginal cost of black coal power (ie fuel + O&M) is typically less than 3 c kWh-1. The 4 c kWh-1 for black coal is the high end of the range  for cost of electricity from a new black coal-fired power station.

·        The capital charge of 20% is unrealistically high. A 25-year plant life and 12% discount rate for example would give an annual capital charge on $40 M capex of $5.2 M. It is also noted that the Barnes model uses FPA's claim of generating 93 jobs to assume this is the number of personnel in the power station.  It is probable that the FPA have considered jobs not only in the power station, but also on the forestry side, plus indirect employment as well. A 30 MW biomass plant should be able to operate on a total wages bill of less than $2 M yr-1.

·        The fuel resource is generally the constraining issue for a biomass plant. If we assume FPA's 280 kt yr-1 and 15% efficiency, this gives a capacity factor for a 30 MW power station of 44%, which is too low to be viable. It is suspected FPA have been working on a much higher conversion efficiency.  For example, a value of 25%  (HHV) would give 73% capacity factor, . However, not too much can be presumed about FPA's calculations on availability of resource throughout the year and power purchase agreements. Using the above assumptions for capital and wages, Barnes assumptions for O&M, and 280 kt yr-1 would give (all at 35% moisture content of fuel):

 

Cost of fuel ($ t-1)

20

30

40

Efficiency (%)

Price of electricity (c kWh-1)

15

12.3

14.7

17.0

20

9.2

11.0

12.8

25

7.4

8.8

10.2

 

 

 

 

References

 

DOE (Department of Energy). http://www.eren.doe.gov/biopower/benefits/be_life_ca.htm

 

DTI (1999).  New and renewable energy prospects in the UK for the 21 st century.  ETSU Report R-122 to the UK Department of Trade and Industry.    Annexe B.

 

NSW MoU (2000). Greenhouse Gas Emissions from Electricity Supplied in NSW: Emissions Workbook.

 

Polglase et al. (1994). Measurement and modelling of carbon storage in a chronosequence of mountain ash forests.  SECV report 63 p.

 

Snowdon et al. (2000). Synthesis of allometrics, review of root biomass and design of future woody biomass sampling strategies. AGO report. 133 p.

 

Go to top of the pageSite created by Adam Schuck
Page last modified: