Is CCS Net-Positive for the Environment?

The short answer is - probably not.

Carbon capture and storage (CCS) is often presented as a unique opportunity to combine significant emissions reductions with the continuing use of carbon-intensive fossil fuel energy sources. In other words, CCS could allow us to continue 'business-as-usual' while minimizing environmental impact. As such, it could serve as a bridge technology from our current system to a future one in which alternative, zero-carbon energy sources are more widely and cheaply available.

That said, if CCS is to have a significant impact on overall emissions, the rate at which full-scale projects must come on-line is quite demanding – the IEA originally estimated a requirement of 100 operating such projects by 2020 if we are to keep global warming below 2 degrees Celsius. According to the Global CCS Institute's latest “Global Status of CCS” report, there are at present a total of 65 'large-scale integrated CCS projects' (LSIP) that are either pending a final investment decision from their proponents, or in current operation (this number is down from 75 in the 2012 report, after thirteen projects were removed and three new ones added). Twelve of these 65 are now in operation, while another 8 have received a final investment decision and are scheduled to come on-line in the next 2 years. Thus it seems that the IEA's original target, while definitely optimistic, is not all that unreasonable.

Project Stage Country Capture (Mtpa) Date Type
Val Verde Operate US 1.3 1972 EOR
Enid Fertlizer Operate US 0.68 1982 EOR
Shute Creek Operate US 7 1986 EOR
Sleipner Operate Norway 0.85 1996 Deep Saline
Weyburn/Midale Operate Canada 3 2000 EOR
In Salah Operate Algeria 2004 Deep Saline
Snøhvit Operate Norway 0.7 2008 Deep Saline
Century Plant Operate US 8.4 2010 EOR
Air Products Operate US 1 2013 EOR
Petrobras Lula Operate Brazil 0.7 2013 EOR
Coffeyville Operate US 1 2013 EOR
Lost Cabin Operate US 1 2013 EOR
Boundary Dam Execute Canada 1 2014 EOR
Illinois Industrial Execute US 1 2014 Deep Saline
Kemper County Execute US 3.5 2014 EOR
Uthmaniyah Execute Saudi Arabia 0.8 2014 EOR
ACTL w/ Agrium Execute Canada 0.5 2015 EOR
Gorgon Execute Australia 3.85 2015 Deep Saline
Quest Execute Canada 1.08 2015 Deep Saline
ACTL w/ North West Execute Canada 1.3 2015 EOR

Where all 65 LSIPs to be operating, the Institute estimates a potential to capture and sequester 122 million tonnes per annum (Mtpa) of carbon dioxide.  The current and soon to be operating projects have a combined potential of 38.6Mtpa (by themselves, however, the currently operating plants have a potential of about 25Mtpa).  Five of these projects have been operating for 10 years or more.

Very few of these projects capture and store CO2 alone, however; most utilize the captured CO2 for enhanced oil recovery (EOR) purposes. The principal behind EOR is to increase the pressure in an oil field already in decline but not too significantly depleted, so as to increase the amount of recoverable resources as well as the rate of production.  The CO2 stays underground after the oil is produced wherein it will remain stored for a very long time, though not all of it. Therefore, whereas a CCS project that stores CO2 for storage purposes alone stores as much as it injects, an EOR project stores somewhat less.  It also adds to the supply of oil, a prime contributor to global energy-related emissions of CO2 in the first place. 

It is also important to bear in mind that this short list of large-scale, integrated projects are by no means the only instances of CO2-based EOR operations currently in existence either. According to the IEA, there were over 100 such projects in existence in 2008, most of which were in North America, producing an additional 300kb/d of oil.1 They also note that the injection of one tonne of CO2 into a suitable reservoir leads to an incremental recovery of between two and three barrels of oil. The EIA's Annual Energy Outlook of the same year suggests that the US potential for EOR could rise from 250kb/d to 1.3mb/d by 2030, 2 and the IEA estimated the world potential to be on the order of 160 to 300 Bbbl. 

Given the potential of CCS projects to increase global production of oil (if used in EOR), it seems therefore a legitimate question to ask whether the CCS projects in current operation and slated to come on line in the next few years will actually have a net-positive impact on reducing energy-related emissions of greenhouse gases (i.e., CO2).  There are, of course, a number of provisions to be made, and numerous uncertainties and variables that prohibit an quick and answer. 

A Quick and Easy Answer

But let's try one anyhow.  If one tonne of injected CO2 leads to 2-3 additional barrels of oil, then the maximum EOR potential of the currently operating CCS projects noted above is only 0.13 – 0.2 mb/d (a small fraction of total world oil supply). According to the US Department of Energy, the GHG content of a barrel of crude oil is 0.43 metric tons of CO2. At that rate, operating EOR-involved CCS projects are having a net impact of somewhere between -3.4Mtpa and 7Mtpa. Adding the deep saline storage projects (assuming no additional associated emissions), the combined net annual impact of currently operating CCS projects would be between -4.9 and 5.4Mtpa. In other words, currently operating CCS projects could be having nearly-equivalent positive or negative impacts in reducing global emissions of CO2, depending on the perfomance rate of the EOR activity (i.e., the incremental production associated with the injection of one tonne of CO2).

Of course, there a numerous problems with this back-of-the-napkin calculation. For one, figures for the amount of emissions embodied in a barrel of oil range considerable – as will be discussed below, there may be more realistic (and lower) figures for an EOR barrel, taking into account the entire life-cycle of the barrel, that differ from conventional crude oil or oilsands derived crudel. Furthermore, this analysis does not consider whether a barrel of oil produced via EOR displaces potentially more environmentally-costly barrels, like those produced in the oil-sands. If this assumption is incorporated, it may be that even though the EOR barrel is not on its own 'net-positive' (i.e., net-negative emissions) it might nonetheless be better than the alternatives. Neither does this approach consider project-specific details that may not align with the IEA's general performance figures - more or less oil could be produced per tonne of CO2 injected, which would change the tally.  Moreover, we have not broached the concept of a recycling rate yet either (the ratio of recycled CO2 - that is, not new CO2 - to the total amount injected.  As a project ages, the recycling rate is prone to increase as more and more of the originally injected new CO2 comes back up.  It can be reinjected, but this supplants the need to bring more new CO2 to the site, thus decreasing the total amount of CO2 stored. 

To address some of these issues, we can look at some previous work done by the Pembina Institute in Alberta as well as a CCS reference case, for which there is much more publicly-available information about operational statistics – the Great Plains Synfuels / Weyburn & Midale CCS project in North Dakota/Saskatchewan.

The Pembina Study

A quite thorough study was conducted early 2013 by the Pembina Institute (contracted by ICO2N) to estimate the net-GHG impact of CO2-EOR projects.  Based on data the authors attained for an EOR project in Alberta, they considered five 'scenarios' to contrast differing perspectives on net impact of such projects.  The first scenario, a reference case, considers a pure CCS project with no EOR.  The second scenario looks at the net emissions associated with EOR from an on-site perspective.  The third looks at the lifecycle emissions of a barrel of EOR, and the fourth and fifth scenarios look at the displacement effects if the EOR barrel replaces either oilsands or conventional crude.

The case study they sue provides the performance ratio (barrels of oil produced per tonne of CO2 equivalent injected) and the recycling ratio (tonnes recycled CO2 injected vs. total volume of CO2 injected), as well as a benchmark value for the on-site energy intensity of producing a barrel of EOR.  The performance rate of this project was 1.1 barrels per tonne of CO2 injected (that is, including both new and recycled CO2), and the recycling rate was 3.45.  At these values, the authors calculate the net per-barrel emissions of a barrel of EOR-produced crude as 0.385 tCO2e, versus 0.43 for oilsands crude and 0.344 for 'average' crude.  It is important to note, however, that this value does not take into consideration the storage of CO2 associated with the barrel of EOR.  At the performance values they found in their case study, they found the net emissions per barrel of EOR - including storage - to be 0.131 tCO2e; still postive, but a lot better than the alternatives. 


In the reference scenario, net storage is effectively 99.9% of injected emissions, since a small amount of emissions are presumed to result from the little energy used to monitor, measure and verify storage.  In the second case, the emissions involved with injecting, producing, recycling and processing the CO2 coupled with other emissions associated with the production of oil, lead net storage to be somewhat less, though still far below break-even – around 70%.   Thus, if a tonne of CO2 is injected, a purely-CCS project will have net emissions of -0.999 tonnes of CO2 equivalent, and an EOR project -0.696 tCO2e.

Things are more interesting in the latter three scenarios.  When looking only at the production and downstream emissions for EOR, the authors find a net amount of emissions generated of 0.5 tCO2e per tonne of CO2 brought to the site.  Take note that this is not the per barrel emissions factor (which, as noted above, is either 0.385 or 0.131, depending on whether you consider storage) - this is net emissions per new CO2 brought to the project.  As such, it us determined to large extent by project characteristics and performance values.

They conducted some sensitivity analysis around some these values; notably, the performance ratio.  The case study they examined achieved only 1.1bbls per tCO2 injected, but more realistic figures for EOR in North America could be higher, the authors note.  As such, they also consider 3bbl and 5bbl per tonne of CO2 brought the site as well.  Given that the largest contributor to net emissions from an EOR project is the consumption of the crude that is produced, more barrels means more net emissions.  At the three performance rates they considerd, the net emissions per barrel (i.e., not per tCO2 injected) of EOR, includng storage, ranges from 0.131, to 0.242, to 0.268 tCO2e / bbl.  On a per-tCO2 injected basis, a performance rate of 3 bbls per tonne of CO2 has net emissions of 2.56 tCO2e (per tCO2 injected). For 5 bbls, the comparable figure is 4.73 tCO2e/tonne brought to the EOR site.

So there is a net positive emissions impact from EOR at all performance values they considered.  When considered to be fully displacing a barrel of oilsands or conventional crude, however, the authors found a barrel of EOR to have a net displacement of -1.175 and -0.834 tCO2e, respectively.  At the performance rates tested in the sensitivity analysis, the net displacement is -1.07tCO2e (-1.98tCO2e for oilsands) per tCO2 injected at a 3bbl performance rate., and -1.31 and -2.84 for conventional and oilsands displacement respectively at the 5bbl ratre..

What do these figures for displacement mean?  They do not mean that by displacing oilsands or conventional crude with EOR we end up with a net negative emissions impact in an absolute sense - after all, the EOR barrel has a positive net emissions impact, irrespective of its displacement effects.    What it does suggest is that, for every tonne of CO2 brought to an EOR site producing at around 3bbls per tCO2 injected, we are doing about that much better than had we not done it at all. Though we still have a net emissions impact of about 2.5 tCO2e from the EOR production, it could have been 3.5tCO2e for conventional crude. 

The Weyburn-Midale CCS Project

Given the base figures we can take from the Pembina study (recognizing of course that they will be at best crude generalizations for the CCS projects we wish to assess below), we need next to consider an operating CCS project case study. The Great Plains Synfuels / Weyburn-Midale EOR project is one operating CCS project for which there is an abundance of documentation.

  Weyburn Midale Pembina
Performance Rate (bbls / tCO2 injected) 1.4 1.6 1.1
Recycling Rate (tCO2 re-injected v. total tCO2 injected) 0.4 0.6 3.45
Net emissions per tCO2 injected (tCO2e) ? ? 0.5
Net emissions per barrel (tCO2e) ? ? 0.131

The project has been in operation since 2000, though pilot testing of CO2 flooding began in 1984 by Shell in the Midale field, and progressed into demonstration project 1992-1999. Since this time, it has been extensively monitored as part of the IEA GHG implementing agreement, the International Energy Agency Greenhouse Gas Weyburn-Midale CO2 Monitoring and Storage. Between 2000 and 2011, there has been an estimate storage of 18MT of CO2. It is estimated that total storage could eventually reach 40 MT.1 The CO2 comes from the Great Plains Synfuel plant in North Dakota, which produces ~13,000 tonnes of CO2 daily, 60% of which is recoverable (and all of that which is captured is utilized in Weyburn/Midale).

The injection rates of new CO2 per field are 6,500 tCO2 per day in Weyburn and 1,250 tCO2 per day in Midale (approximately 2.8Mtpa per year), and roughly 13,000 tCO2e and 1,650 tCO2e, respectively, including recycled CO2. This rate of injection has increased production by 18k bbl/d at Weyburn and 2.3k bbl/d at Midale. The total projected additional production over the lifetime of the project is 215 Mbbl for both fields combined.2  

A very basic calculation, using the US DOE figure for emissions per barrel of oil, might go as follows:

  • Total storage = 40 MT

  • Total production = 215 Mbbl

  • Total additional emissions = 215 Mbbl * 0.43 tCO2e = 92.45 MT

  • Net emissions = 92.45 MT – 40 MT = 52.45 MT CO2e

Therefore, over the lifetime of the projection, Weyburn/Midale could be expected to have a net negative impact on the environment of 50.45MT of CO2 over its lifetime (not considering if it is displacing other types of oil).  Let's consider a few scenarios, based on the Pembina study's data. 

Weyburn-Midale Emissions Scenarios

First off, let's consider if there had been no CO2 capture in the first place (the 'upstream' benefits of EOR where not considered by Pembina, but they are worth noting). In this case, the Great Plains Synfuels plant would continue to produce its ~13,000 tonnes of CO2 a day, but without capturing the 60% it does now. That means the 7,750 tCO2 that Weyburn/Midale brings to the site each day would be release into the atmosphere. Over the 30 year lifetime of the project, the total emissions that could have been captured and stored (maximally speaking) could reach almost 85 MT. But Weyburn/Midale can only store so much, so our base case is essentially 40 MT stored (which is the estimated storage capacity of the fields noted above). 

Of course, the production of oil means that there will be some emissions associated with the project.  Recall that the Pembina study found that the net emissions resulting from one tonne of CO2 brought to an EOR site, including downstream emissions associated with the resulting barrel of oil, was 0.5 tCO2e (assuming 1.1 bbls produced per tonne of CO2 injected). At those rates, given that Weyburn/Midale brings a total of 7,750 tonnes of CO2 to the fields daily, total net annual emissions from the resulting production would be about 1.4 MT CO2e – 42 MT over the life of the project.  So we've gone from storing 40 MT to producing an equivalent amount.

As indicated above, however, the actual performance rates in Weyburn / Midale (that is, incremental oil produced per unit of CO2 injected) are about 1.4 bbl/tCO2e injected in Weyburn and 1.6 bbl/tCO2e at Midale, which would suggest slightly higher net emissions per tonne-injected. The recycle rates for Weyburn / Midale are considerably lower, however – 0.5 and 0.3 respectively, compared to 3.45 in Pembina's baseline case. This means that the net per-barrel emissions at Weyburn-Midale are likely somewhat lower than they are in the Pembina case, since less CO2 is being recycled and more new CO2 is being used (depending, however, on the amount of electricity used in production and the manner in which it is generated).  

These rates will likely change over time, though it is difficult to anticipate how (the recycling rate will probably increase overtime, though if the area into which CO2 is being injected is expanded the recycling rate would decline again; also, it is difficult to estimate how incremental production may change over time). Furthermore, there is no public data about the energy use at Weyburn-Midale, which is the single biggest component of on-site emissions involved in EOR activities. Without this information, it is difficult to come to a more exact figure for the net emissions stemming from Weyburn-Midale now, or in the future. 

We have not yet considered the displacement effects the EOR-produced barrels may have. According to the Pembina study, at a performance rate of 1.1bbls/tCO2 injected, their case study has a net life-cycle displacement effect -1.175 tCO2e, if the barrel displaced is from the oilsands, and -0.834 tCO2e if it is 'average' conventional crude. Extrapolating on those figures, we might assume that if Weyburn/Midale's performance remained steady over its life-time of 30 years and all of the additional oil produced supplanted dirtier sources, the net displacement impact would be between -70.8 and -99.7 MT.  That said, the Pembina authors note that over the long-term, EOR production is more likely to add to the total supply, rather than displace any of it.  Without much more in-depth ecnomic modeling and more accurate and specific data for the key parameters, however, arriving at realistic estimates of displacement and net-emissions is not possible.

By extension...

There are 8 other operating LSIPs that are being used in enhanced oil recovery projects. These are: Val Verde; Enid Fertilizer; Shute Creek; Century Plant; Air Products; Petrobras Lula Oil Field; Coffeyville and Lost Cabin. All except one of these are in the US. There is no publicly available data about the performance or recycling rates of these projects, the energy consumption associated, and the projected life-time of the projects are hard to come by as well. Making anything more than an extremely crude assessment based the project lifetime of Weyburn/Midale, using the Pembina figures for another, undisclosed EOR case study is thereby not possible.

But let's do so anyhow.  The table below, summarizes the operating and soon-to-be operating CCS projects for their storage and net-emissions impacts, based on the assumption that all of them have equivalent operating characteristics as the Pembina study (i.e., 0.5 tCO2e per tCO2 injected for EOR and 0.999 storage for CCS only). 

Summary No. of Plants EOR Deep Saline
Operate 12 9 3
Execute 8 5 3
Total 20 14 6
Potential Storage, Mtpa EOR Deep Saline
Operate 25.63 24.08 1.6
Execute 13.03 7.1 5.9
Total 38.66 31.18 7.5
Potential Net Emissions, Mtpa EOR Deep Saline
Operate 10.5 12.0 -1.5
Execute -2.4 3.6 -5.9
Total 8.1 15.6 -7.5

Absent any displacement affects, we can see that there could be a net positive contribution of these CCS projects of 8 MT of CO2 per annum.  If each operates as long as Weyburn-Midale can be expected to, that equals 240 MT of CO2 over and above what would be stored. That being said, not capturing and storing the 38 MT per year that these projects aim to would produce a whole lot more net emissions, so perhaps the benefit is real. 

Weyburn-Midale is expected to store 40 MT and produce 215 Mbbl of oil, though it brings approximately 2.8 MT per year to the site (84 MT over 30 years).  That's about 5 million barrels of oil per megatonne of CO2 stored, but 2.6 million barrels per megatonne of CO2 brought to the site.  Assuming similar values for the other CCS projects (a very questionable assumption, to be sure), thats about 2.4 billion extra barrels of oil produced by the CCS-EOR projects considered here.   At the net-emissions per-barrel value of 0.131 tCO2e found by Pembina, that's an additional 313 MT of CO2 over thirty years for the EOR projects alone, and 88.6 MT CO2 when taking out the proportion stored by deep saline projects. .  

So, in conclusion, CCS when used for EOR is going to have a net contribution to global GHGs above and beyond what is stored during the process.  The assumptions used in this short analysis are way to crude to be taken too seriously, though they do suggest possible extreme-case scenarios in which CCS ends up producing a very signficant amount of CO2. 

1International Energy Agency, World Energy Outlook 2008, 213.

2Energy Information Administration, Annual Energy Outlook 2008: With Projections to 2030, 36.

3Energy Information Administration, Annual Energy Outlook 2013.

1Whittaker et al., “A Decade of CO2 Injection into Depleting Oil Fields,” 6071.

2Ibid., 6070