Inefficient power generation as an optimal route to negative emissions via BECCS?

Three distinct scenarios were considered in this study, and are described below.

• Scenario 1 (Sc1) considers the BECCS plants to operate in response to a demand for electricity—what would be traditionally considered to be a load-following manner. They burn biomass to produce power in response to a market demand, and remove CO₂ from the atmosphere as a consequence.
• Scenario 2 (Sc2) considers the BECCS plants to operate in a baseload fashion, constantly removing CO₂ from the atmosphere, but dispatching electricity only when there is a demand. When surplus electricity is generated, it is 'spilled'⁶, and no payment is assumed. This is a relatively common response within energy systems with periods of oversupply [26]
• Scenario 3 (Sc3) considers the BECCS plants to operate in a baseload fashion, constantly removing CO₂ from the atmosphere, but dispatching electricity only when there is a demand. When surplus electricity is generated, it is directed to an electrolysis plant to produce hydrogen which is, in turn, sold to provide a heating service. The value of this hydrogen is indexed off of current methane market values on a mass-based energy content. It is also conceivable that this H₂ could also be used as a transport fuel or as an industrial feedstock. However, for this study, we have focused solely on the power-to-gas model.

The purpose of this model is to calculate the net present value (NPV) and internal rate of return (IRR) of a BECCS plant and subsequently evaluate the number of plants, NP, required to meet a given negative emissions target, NT, and the cost per tonne of CO₂ removed, CPTR. The NPV of a given investment is obtained from

$$\begin{equation} \text{NPV}(n,r) = \sum\limits_{N = 0}^N {\frac{{\text{CF}_n }}{{(1 + r)^n }}} \end{equation} \tag{ 1 }$$

where n and r are the year of a given cash flow, CF, and the discount rate applied to that CF, respectively. The IRR of a given investment is, formally, that rate r which results in an NPV of zero, obtained by solving the following expression for r

$$\begin{equation} \text{IRR}(n,r) = \sum\limits_{n = 0}^N {\frac{{\text{CF}_n }}{{(1 + r)^n }} = 0.} \end{equation} \tag{ 2 }$$

The NP required to meet NT is obtained from

$$\begin{equation} \text{NP} = \frac{{\text{NT}}}{{\text{CS}\text{.MWhYr}_O .10^{ - 6} }} \end{equation} \tag{ 3 }$$

where CS is the quantity of CO₂ sequestered per MWh and MWhYr_O is the number of MWh per year and is in turn calculated via

$$\begin{equation} \text{MWhYr}_O = \text{NPC}\text{.LF}_O .8760 \end{equation} \tag{ 4 }$$

where NPC is the nameplate capacity of the BECCS plant and LF_O is the average operating load factor of the BECCS plant. Similarly, the levelised cost per tonne of CO₂ removed is calculated via

$$\begin{equation} \text{CPTR} = \frac{{\text{EAC}.10^6 + \text{MC}\text{.MWhYr}_O }}{{\text{CS}\text{.MWhYr}_O }}. \end{equation} \tag{ 5 }$$

Here, MC is the marginal cost of operating the BECCS plant and is taken to be the sum of the fuel costs, FC, CO₂ transport and storage costs, CTS, and variable operating and maintenance costs, VOM;

$$\begin{equation} \text{MC} = \text{FC} + \text{VOM} + \text{CTS}. \end{equation} \tag{ 6 }$$

In this study, we consider that biomass combustion is carbon neutral, and therefore there is no cost imposed for the emission of 'bio-CO₂'. Thus, the equivalent annual cost, EAC, of the BECCS plant is calculated via

$$\begin{equation} \text{EAC} = \frac{{\text{CPX}}}{{A_{n,r} }} \end{equation} \tag{ 7 }$$

where CPX is the capital cost of the BECCS plant. In this study, CPX is estimated via a regression of data from IECM [27] in terms of higher heating value efficiency, η_HHV and is given by

$$\begin{equation} \text{CPX} = \frac{{932.7\eta _{\text{HHV}} + 1153.2}}{{\text{CCF}}}. \end{equation} \tag{ 8 }$$

Here, CCF is a currency conversion between GBP and USD, taken to be £1 = $1.3 in this study. The annuity factor, A_n,r in equation (7) is given by

$$\begin{equation} A_{n,r} = \frac{{1 - \tfrac{1}{{(1 + r)^n }}}}{r} \end{equation} \tag{ 9 }$$

The cash flow of the BECCS plant in any year n is taken to be the sum of the revenue streams associated with operating the plant, R_O and with dispatching electricity, R_D:

$$\begin{equation} \text{CF} = \left( {R_O .\text{MWhYr}_O + R_D .\text{MWhYr}_D } \right)10^{ - 6} . \end{equation} \tag{ 10 }$$

In equation (10), we introduce the distinction between scenario 1 and scenarios 2–3 by distinguishing between the number of MWh per year for which the plant is operating, MWhY r_O and that for which it is dispatching electricity MWhYr_D. Each of these quantities are defined as follows:

$$\begin{equation} \text{MWhYr}_O = \text{NPC}.\text{LF}_O .8760 \end{equation} \tag{ 11 }$$

and

$$\begin{equation} \text{MWhYr}_D = \text{NPC}.\text{LF}_D .8760 \end{equation} \tag{ 12 }$$

where NPC is the nameplate capacity of the BECCS plant and LF_O and LF_D correspond to the operating and dispatching load factors, respectively. The R_O is simply the difference between income derived from removing CO₂ from the atmosphere and the marginal cost of BECCS plant operation, MC

$$\begin{equation} \text{R}_O = \text{CP}.\text{CS} - \text{MC} \end{equation} \tag{ 13 }$$

where CP is the payment per tonne of CO₂ removed from the atmosphere and CS is the quantity of CO₂ removed from the atmosphere per MWh generated. However, it must also be recognised that whilst we consider the bio-CO₂ to be carbon neutral, this does not mean that one should receive a negative emissions credit for all of the combustion CO₂ sequestered. This amount should properly be reduced by the amount of CO₂ emitted throughout the biomass supply chain, i.e. the biomass carbon footprint, BCF. Thus, the value of CS is given by:

$$\begin{equation} \text{CS} = \text{CI}.\text{CCS} - \frac{{\text{BCF}}}{{\eta _{\text{HHV}} }} \end{equation} \tag{ 14 }$$

where CI is the carbon intensity of the BECCS plant and CCS is the fraction of CO₂ produced from fuel combustion that is captured and sequestered, taken to be 90% in this study. The plant CI is, of course, a function of the plant efficiency and the carbon content of the fuel, CI_F,

$$\begin{equation} \text{CI} = \frac{{\text{CI}_F }}{{\eta _{\text{HHV}} }} \end{equation} \tag{ 15 }$$

and

$$\begin{equation} \text{CI}_F = \frac{{C\tfrac{{44}}{{12}}}}{{\rho _F^E .2.777 \times 10^{ - 7} }}. \end{equation} \tag{ 16 }$$

In equation (16), C is the carbon content of the biomass, $\rho_F^E$ is the energy density of the biomass. In this study, single representative values of C and $\rho_F^E$ are used. Obviously, it is possible to find a range of values for both quantities. In practice, however, the carbon content of biomass is typically in the range 47%–49% and it is therefore conceivable that a single average value might be adopted for ease of regulation. Finally, the factor of 2.777 × 10⁻⁷ is to convert MJ to MWh and kg to tonnes⁷. The revenue derived from dispatching electricity is simply

$$\begin{equation} R_D = \text{EP}.\text{MWhYr}_D \end{equation} \tag{ 17 }$$

where EP is the electricity price and MWhYr_D is as defined in equation (12). The connection between plant efficiency and load factor has been neglected in this study in order to preserve efficiency as an independent variable. In order to calculate the MC, we need to calculate FC, CTS and VOM. The fuel costs are calculated via

$$\begin{equation} \text{FC} = \frac{{\text{FC}_M .1 \times 10^{ - 3} }}{{\eta _{\text{HHV}} .\rho _F^E .2.777 \times 10^{ - 4} }} \end{equation} \tag{ 18 }$$

where FC_M is the mass-based fuel cost, taken to be £50 t⁻¹ in this study, and is line with the costs reported in Seikfar et al [28]. This is a representative value, with biomass prices having the potential to vary substantially in practice. Lastly, the cost associated with CO₂ transport and storage, CTS is given by

$$\begin{equation} \text{CTS} = \text{CD}_M .\text{CS} \end{equation} \tag{ 19 }$$

where CD_M is the cost per tonne for CO₂ transport and storage, assumed to be £20/t_CO2 in this study, and CS is as defined in equation (14). In order to evaluate the role and value of hydrogen production, we first quantify the amount of electricity which was produced but not dispatched, Q_S via

$$\begin{equation} \text{Q}_S = \text{MWhYr}_P - \text{MWhYr}_D \end{equation} \tag{ 20 }$$

and then the amount of hydrogen produced, HP, is found directly

$$\begin{equation} \text{HP} = \frac{{Q_S .1 \times 10^3 }}{{\text{HC}}}, \end{equation} \tag{ 21 }$$

where HC is the power required to produce 1 kg of H₂, taken to be 53.4 kWh kg⁻¹ in this study [29]. We then proceed to calculate the number of electrolyser units required to produce this much H₂, NE,

$$\begin{equation} \text{NE} = \frac{{\text{HP}}}{{\text{EC}.\text{LF}_O .8760}} \end{equation} \tag{ 22 }$$

where EC is the capacity of an individual electrolyser. An expression for the specific capital cost of each electrolyser, SPCPX, as a function of EC was estimated from published data [29] and is

$$\begin{equation} \text{SPCPX} = \frac{{18.\text{EC}^{0.6} }}{{\text{CCF}}} \end{equation} \tag{ 23 }$$

where SPCPX has units of k£. Finally, the total capital cost of the electrolysis plant, ECPX, in M£, is estimated from

$$\begin{equation} \text{ECPX} = \text{NE}.\text{SPCPX}.1 \times 10^{ - 3} . \end{equation} \tag{ 24 }$$

Therefore, for Sc3, the term ECPX was added to the capital cost of the BECCS plant defined by equation (8). Where hydrogen was produced and sold, it was assumed to displace natural gas (primarily CH₄) for heating and therefore its value was indexed to that of CH₄ on a heating value basis [30]. For this exercise the CH₄ price was assumed to be £25 MWh⁻¹ [31] and H₂ was assumed to have an energy density of 39 kWh kg⁻¹, and therefore H₂ could be sold for approximately £1 kg⁻¹. Again, this was simply added to equation (13) in order to calculate the revenue under Sc3.

The construction period of a CCS plant is typically estimated to be between 4–6 yr [32], and an estimate of 5 yr was used in this study. Similarly, whilst it is common for large industrial facilities to take up to four years to achieve design point operation [33], this nuance was not considered here - we assumed that the BECCS plant was capable of design point operation as soon as construction was completed. A technical project lifetime of 50 yr (including construction) was assumed. For project financing, a debt-to-equity ratio of 70/30 was assumed, with debt assumed to be available at 5% and equity valued at 8%, yielding a weighted average cost of capital (WACC) of approximately 6%, this is in line with other estimates [34]. Future cash flows were discounted at a rate of 5%. The biomass fuel was considered to be woodchips and a conservative cost of £50 t⁻¹ was assumed [28]. This biomass was assumed to have an energy density of 18.2 MJ kg⁻¹ with 5% moisture and to be 47.5 wt% carbon⁸. In line with common practice [35, 36], 90% CO₂ capture was assumed in all cases. CO₂ transport and storage costs of £20/t_CO2 and variable operating and maintenance costs of £4.38 MWh⁻¹ were assumed [37]. Lastly, an average wholesale price of electricity of £80 MWh⁻¹ was assumed [24]. On this basis, and assuming a reference BECCS plant with η_HHV = 42% and a load factor of 85%, a central value of £75/t_CO2 as a payment for the CO₂ removed from the atmosphere was chosen as it was found to provide an IRR of 12%⁹, noting that this price is a linear function of both biomass cost and electricity price. For comparison, in the absence of this payment for removing CO₂ from the atmosphere, in order to achieve an IRR of 12%, the BECCS plant would need to receive a payment of approximately £130 MWh⁻¹ of electricity dispatched, still assuming an 85% load factor.

We begin our analysis by considering a discounted cash flow analysis (DCF) of a BECCS facility. In figure 1, we present a discounted cash flow analysis of both the high- and low-efficiency BECCS facilities for each of scenarios 1–3. For this illustrative example, we consider the limits of our range, i.e. η_HHV of 30 and 45%, respectively.

Zoom In Zoom Out Reset image size

In all cases, the inefficient BECCS facility is more profitable than its more efficient counterpart, assuming similar operating patterns. In the case of Sc1, the inefficient plant has an NPV of approximately £1982 M, relative to that of the efficient plant at approximately £1469 M. In Sc2, this increases by 16% and 13%, respectively. The addition of hydrogen production in Sc3 has a further positive effect, adding a 9% and 12% to the NPV, respectively. A detailed breakdown of the un-discounted cash flow is presented in figure 2 for clarity.

Zoom In Zoom Out Reset image size

As can be observed, the primary cost incurred over the lifetime of the BECCS plant, regardless of scenario, is fuel, followed by the cost associated with CO₂ transport and storage. The primary source of income in Sc1 is the sale of electricity, closely followed by the payment received for removing CO₂ from the atmosphere, but once the BECCS plant operates at constant capacity, regardless of electricity dispatch, this payment substantially increases, substantially improving the total positive cash flow, and in the case of the less efficient option, reversing entirely. The income associated with selling hydrogen is observed to be a non-negligible but is a distant third under this scenario. Here, we considered that the H₂ was sold in order to displace natural gas for space heating, and as such commanded the relatively low price of £1/kg_H2.

An important element which was not considered in this analysis is any potential income arising from CO₂ emissions that are avoided as a result of utilising of CH₄. For the BECCS plant operating with a 60% dispatch factor in Sc3, this is equivalent to producing approximately 20 506 t_H2/yr, which displaces approximately 55 680 t_CH4/yr and thus avoids the emission of 0.15 Mt_CO2/yr. Assuming that a similar payment was available for CO₂ avoided as for CO₂ removed, this would correspond to an additional income on the order of £11 M yr⁻¹.

However, whilst it is evident that the income derived from hydrogen production is potentially important, it does not qualitatively affect the results of this work, and will therefore not be discussed further in this analysis. The key observation at this point is that the less efficient plants will of necessity burn more biomass over the plant lifetime, and therefore has the potential to remove substantially more CO₂ from the atmosphere than the more efficient plant.

Building on the previous discussion, the NPV of a given plant is clearly a function of the power plant efficiency, the average load factor and the income derived from removing CO₂ from the atmosphere. This was explored in depth, and the associated results are presented in figure 3.

Zoom In Zoom Out Reset image size

As can be qualitatively observed from figure 3, there is a complex tradeoff between efficiency, load factor and CO₂ price. At low CO₂ prices, as in figures 3(a) and (d), the dispatch load factor exerts a first order effect on the NPV. As the CO₂ price increases, the influence of dispatch load factor on NPV is reduced. In all cases, the less efficient plants are more profitable at a given load factor than the more efficient plant—this trend may be qualitatively observed from figures 3(b), (c), (e) and (f).

These trends are somewhat more evident when the viability of this investment is expressed in terms of the internal rate of return, as in figure 4.

Zoom In Zoom Out Reset image size

As can be observed from figure 4, IRR is inversely proportional to BECCS plant efficiency. Tellingly, for a given load-factor, the less efficient plant always has a greater IRR than the more efficient plant.

These results can be conveniently summarised in terms of the total amount of CO₂ removed from the atmosphere per year per plant, the number of 500 MW units required to meet a target of −50 Mt_CO2 and the levelised cost per tonne of CO₂ removed from the atmosphere over the lifetime of the BECCS facility. This target was chosen as it is in line with potential UK targets [38]. These metrics are illustrated in figure 5.

Zoom In Zoom Out Reset image size

Several points are immediately apparent from figure 5. Firstly, the distinction between Sc1 and Sc2 is substantially more apparent here than in the previous graphics. Firstly, the less efficient plants remove substantially more CO₂ per year than the more efficient plants. This is the chief motivation for operating the less efficient BECCS plants in baseload fashion - it results in an increase in the total amount of CO₂ removed from the atmosphere per unit per year; approximately 2.7 Mt_CO2/yr to approximately 4.1 Mt_CO2/yr for the most and least efficient plants, respectively. This, in turn, substantially reduces the number of units required to achieve the aforementioned target of −50 Mt_CO2/yr and thus a reduced marginal abatement cost.

The previous analysis was performed from the perspective of a BECCS power plant, neglecting the biomass supply chain, using the simplifying assumption that the biomass delivered to the power plant was carbon neutral. This is, of course, not true, and CO₂ emissions associated with biomass cultivation, harvesting, processing and transport can be relatively high [39, 40], and would, if included in the carbon balance on the system, decrease the carbon negativity of the BECCS power plant, or, if sufficiently great, lead to the BECCS facility being a net emitter of CO₂.

The UK Bioenergy Strategy [41] identifies a value of 285 kg_CO2eq per MWh of bioelectricity as the upper limit of GHG emissions embedded in the supply chain for that bioenergy to be considered sustainable. However, owing to the uncertainty associated with their quantification, indirect land use change (ILUC) emissions are not included within this definition. For reference, Drax reports [42] a value of 36 g_CO2eq per MJ of electricity generated, or between 39–58 kg_CO2eq per MWh of primary bioenergy, as a function of the efficiency of the conversion process.

Thus, in order to evaluate the impact of including embodied GHG emissions on the foregoing results, we evaluated the impact of using a fuel with a biomass carbon footprint (BCF) in the range 0–300 kg_CO2eq per MWh. As before, the performance indicators were NPV, IRR, cost per ton of CO₂ removed (CPTR) and annual CO₂ sequestered (CS) of a BECCS project, and their variation with the system power generation efficiency, under the previously described operating scenarios. For this analysis, the dispatch load factor, electricity price and negative emission credit were set to 60%, £80 MWh⁻¹ and £50/t_CO2, respectively. The results of this analysis are presented in figures 6 and 7.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In figures 6(a) and (b), i.e. Sc1, for all values of BCF less than 135 kg_CO2eq per MWh of primary bioenergy delivered, the less efficient power plants are uniformly more profitable than their more efficient counterparts. For reference, this is approximately 2.8 times greater than the value reported by Drax. However, above this value, the trend begins to reverse, and the more efficient units become more profitable. What is being observed is that, as the system efficiency decreases, the increase in revenue associated with the negative emission credit becomes too low to compensate for the decrease in the electricity revenue, resulting in an overall lower NPV and IRR for less efficient systems.

In Sc2, the point at which this transition occurs is reduced to 90 kg_CO2eq per MWh. Here, recall that the plant is operating in baseload fashion, i.e. sequestering CO₂ at an 85% capacity factor but dispatching electricity at 60% of nameplate capacity. Therefore, the revenue associated with the sale of electricity is reduced relative to the short run marginal cost, and thus overall profitability is reduced.

An evaluation of the biomass carbon footprint on the cost per tonne of CO₂ removed from the atmosphere and the amount of CO₂ removed per year is presented in figure 7.

As can be observed from figure 7, at low carbon footprint values, efficiency has a first order impact on both the cost per tonne of CO₂ removed (CPTR) and also the amount of CO₂ removed from the atmosphere (CR) of a given BECCS facility. However, as the biomass carbon footprint increases, the importance of plant efficiency decreases. For a biomass carbon footprint greater than 270 kg_CO2eq per MWh, the cost of removing one ton of CO₂ is observed to significantly increase. Specifically from £640/t_CO2 at 270 kg_CO2eq per MWh to £2500/t_CO2 at 300 kg_CO2eq per MWh for a 45% efficient power plant. This can be observed in the figures 7(a) and (c) as an abrupt transition from blue to green. This can be elucidated by inspection of figures 7 (b) and (d). Initially, the less efficient plants remove significantly more CO₂ from the atmosphere than the more efficient plants. However, as the biomass carbon footprint increases, the amount of CO₂ removed per year decreases, and beyond 270 kg_CO2eq per MWh, the impact of power plant efficiency becomes negligible relative to that of the embodied carbon footprint of the biomass.

This serves to reinforce the original hypothesis of this study that for a given system, power plants that are less efficient at converting primary bioenergy into bioelectricity, and thus are commensurately lower in capital cost, will consistently remove more CO₂ from the atmosphere per year and at a lower cost per tonne removed than their more efficient counter parts.

The only point where this hypothesis breaks down is where the embodied carbon in the bioenergy is sufficiently large so as to render the impact of power plant efficiency negligible and thus the dominant factor is the power plant capital cost, and the sale of electricity becomes more valuable than the removal of CO₂ from the atmosphere.

Given that the purpose of building and operating the BECCS facilities is arguably to achieve least cost atmospheric CO₂ removal, then what might initially be considered to be a sub-optimal choice—building BECCS facilities which are inefficient at converting biomass to electricity and operating them in baseload fashion—may actually be the cost optimal route to removing CO₂ from the atmosphere.