Skip to main content
European Commission logo print header

Large-scale CCS Transportation infrastructure in Europe

Final Report Summary - COCATE (Large-scale CCS transportation infrastructure in Europe)

Executive summary:

The purpose of the COCATE project was to explore the possible implementation of a shared transport system for flue gases from industries and for carbon dioxide (CO2), once extracted from the flue gas.

From the study it appears that sharing the capture can be a good solution for small emitters even if there are no easy rules to choose one collecting solution more than another and if it is difficult to identify roughly.

Data collected in Le Havre industrial zone were the basis of the different studies performed on the three necessary networks:

- Low pressure: flue gases (and alternative)
It was shown that the composition and volume of the flue gases could pose some difficulties. Indeed it implied large pipe diameters, difficult routing, and using stainless steel or coating as material construction to cope with corrosion and possible erosion which implied higher cost.

- Medium / high pressure: CO2 from capture to Hubs
The impact of some impurities on the transport was studied. CO2+N2O and CO2+NO pressure-composition and pressure-density phase diagrams were performed at different temperatures in the range (-20, 20 degrees of Celsius) using Monte Carlo molecular simulation and cubic equations of state (EOS). The degree of conversion in the NO+N2O2 equilibrium was investigated. Thermal conductivity of CO2+N2, CO2+O2 and CO2+H2S mixtures was defined at 130 bar and 150 bar for two temperatures: 15 and 20 degrees of Celsius by molecular simulation and compared to values obtained with selected EOS. Failures were simulated to identify their potential impact on the design.

- High pressure: export system to final storage (pipe, ship)
This network was not the focus of this project, but was integrated in the economics (economic CO2 network optimisation model) and risk analysis.

Risk analysis associated with the different design showed that further studies will be needed if it was decided to implement a network such as the one developed in COCATE. It was also shown that the business model that would allow a common capture and transport network for such industrial areas, is complex and could potentially slow down the development of such equipment.

In order to facilitate the decision making process and increase chances of implementation a decision making tree was developed. This tree, based on a real case, can help other similar industrial areas (400 km2, 10 Mt CO2 / y) to develop their own shared collection network.

Project context and objectives:

All ongoing projects on CO2 transport are focusing on the origin of the CO2 coming from power plants having their own CO2 capture process. No single project considers multiple different smaller sources. The technical problems involved in this are different due to the multitude combination of fumes and CO2 mixtures; further, the need of a low pressure pooling network adds additional complexity to the problem.

According to this diagnostic, this project aims at exploring the challenges generated by a large-scale deployment of CO2 capture and storage (CCS) infrastructure connecting different small CO2 sources located in the same industrial area, to a final storage site.

Three networks have to be considered, the first one collecting the fumes is a low pressure high-temperature network transporting a very corrosive mixture. The second one, after the CO2 is captured transport the flow up to the conditioning unit. The third one is composed of pipelines, intermediate storage and / or boats and transports the CO2 gathered to the storage sites (the link to the offshore final pipeline in a harbour is our final boundary in this project). Different gaps were identified impacting the network life (impurities, network management) or the safety and acceptance (risk analysis including failure modes, guidance on acceptance).

At the present time, some geographical zones are already considering grouping the fumes collected from different emitters in order to reduce treatment costs. This is the case for Le Havre (France). Rotterdam (Netherlands) has a strong initiative for collecting and storing the CO2 emitted in the harbour area. In order to base the study on real constraints those two areas have been chosen to represent a group of industrial pooling the capture plants and a hub through which the CO2 collected will be definitively store.

Le Havre industrial area emits today more than 14 Mt CO2 per year. For the study and based on 2009 projections, a total mass of CO2 emission of approximately 20 Mt per year is considered.

The mixture of industrial sources result in a great variety of CO2 streams characteristics: daily rates, annual rates, CO2 content, type and concentration of other gases and components, volume, pressure, temperature.

To achieve the global objective of the project the following objectives were identified:

(a) define the overall collecting and transport network from sources to sinks in order to get data for all the technical and research studies;
(b) perform thermodynamic studies of CO2 + impurities mixture and study some flow parameters such are thermal conductivity;
(c) make some corrosion tests on material, check the software evaluations and look at prevention solutions such as internal coatings;
(d) estimate the hydrodynamics behavior in the collecting network to identify possible design impossibilities / solutions;
(e) define, quantify and analyse (HSE) risks related to the overall collecting and transport chain of CO2 from sources to sinks, in doing so: regarding both industrial and external safety issues, focusing on the present uncertainties in modelling failure scenarios and effects;
(f) identify key economic and regulatory drivers for regional pooling and network of transport of captured CO2 streams from large CO2 projects (micro-economics);
(g) define business models and investment strategy relevant for large scale transport of CO2 connecting clusters of emitters to various sinks (macro economics);
(h) develop an economic CO2 network optimisation model, with application to Le Havre case.

This real application is of great benefit for other industrial areas and for the ongoing CCS South African actions more focused at the present time on capture and storage than on transport solutions (currently solutions based on trucks are retained in the studies).

A strong effort to keep each partner, including the Commission, fully informed about the project status and the technical issues and with parties outside the consortium is the key success of this project. During the last year a focused is applied on the public dissemination of COCATE results.

Project results:

In order to explore the challenges generated by a large scale deployment of CCS infrastructure connecting different small CO2 sources located in the same industrial area, to a final storage site, a real case is taken to furnish the necessary data to COCATE study. Le Havre is identified to be the real case for the pooling strategy and Rotterdam is considered to be our connection to the final storage.

13 emitters are considered which represents 95% of the total emissions issued from the Le Havre / Port Jérôme industrial basin. For these 13 industries, 85 point sources have been registered having emissions from 2 000 to 2 500 000 tCO2 / y. Most of the streams are coming out of combustion processes but it is noteworthy that the characteristics of the flue gas streams coming from these sources are quite different. Some sources have specific characteristics and need to be analysed separately.

Some treatments after combustion are already in place on some of the units but not on all of them.

As we are dealing with existing industrial sites it is decided to consider only a post-combustion technology for capture: absorption of CO2 by an amine solvent. This choice is validated by the industrial partners involved in the project.

This technology offers several clustering options:

- Common capture centre: flue gas coordinated transport up to a common capture centre (several capture centres can exist on an industrial basin depending on the localisation of the industries).
- Common regeneration centre: CO2 is captured on site (absorption column next to the stack), the CO2 rich solvent is transported up to a common regeneration centre and lean solvent is sent back to the absorption column.
- Common conditioning centre: CO2 is captured and solvent is regenerated on site next to the stack, and CO2 is transported in the gaseous phase through a mutualised network to a hub where it is further conditioned before it is exported to the storage sites.
- Common 'energy' centre: all the infrastructures necessary for capture and conditioning of CO2 are next to the stack, only the utilities (steam and electricity) necessary to this process are produced by a common energy centre before being distributed to the different capture units.

Combinations of those clustering options are possible, keeping in mind that each solution has specific advantages and drawbacks (e.g. common capture centres allow small emitters to benefit from the economies of scale of the capture process).

The highest degree of clustering is achieved through the common capture centre option, which is chosen as the starting point for the COCATE project. Data analysis leads to five pooling centres within the industrial area for this preliminary scenario. For some of them 2 alternative solutions are possible.

Once the pooling centres are defined, studies revolve around three networks:

- Low pressure: flue gases (and alternative).
- Medium pressure: CO2 from capture to Hubs.
- High pressure: export system to final storage (pipe, ship).

The low pressure network mostly transports very corrosive mixtures, therefore studies focused on corrosion aspects. For the medium and high pressure ones even if most of the impurities are removed attention was geared towards thermodynamics estimations, safety and network management.

Based on the results for the different network a preliminary implementation strategy is proposed for the Le Havre / Port Jérôme area. It is presented into more details in §1.3.4.

Finally, to insure that developments made through COCATE will benefit other industrial clusters, industrial areas are identified elsewhere in Europe and South Africa and recommendations on a generic strategy of deployment is made (see §1.3.4).

1.3.1 Low pressure: flue gases (and alternative)

From the real case it raises that the conditions supported by such networks are critical:
(a) temperatures range from 70 to 470degrees of Celsius (most of the temperatures from 100 to 250 degrees of Celsius);
(b) pressures are below 1.1 bara;
(c) impurities are present (H2O up to 30 %, O2 up to 10 %, NOx up to 0.04 %, SOx up to 0.6 %, CO up to 0.4 %, other possible components such as VOC, N2O, CH4, HF, HCl, Ar, Heavy Metals). Presence of particulate matter up to 200 mg/Nm3 no information on their nature.

Studying the integrity of this part of the network, potential erosion due the very high velocities in these pipes appears. If solid particles are present in the flux then these phenomena can become critical for the design and solutions to remove directly the particles at the source must be considered.

Particular attention has to be paid to the network management studies as high volume of fume have to be accounted for and impact directly the network design. Then different options are studied including parallel lines, pressurised lines or more 'exotic' solution such as the separation of the absorption and regenerator columns of the capture plant with an amine line between both. After having checked the technical feasibility of each solution an economic benchmark is done. When very large amounts of flue gases or huge low CO2-concentrated sources have to be treated and transported the amine line solution appears to be the best. Careful consideration of the thermodynamics and chemical reactions associated with leaks from the amine solution circulation system indicate insignificant safety risks due to the resistance of the amine solution to de-gassing at ambient temperatures.

On this network the CO2 threshold cost range in 31 - 39 EUR / tCO2.

Moreover, the benefit of pooling emissions in terms of costs for small emitters was also assessed. As for example in pooling #4, pooling emissions and capturing CO2 at a common site compared to the stand alone capture option allowed:

(a) a reduction of 20 % of the cost discounted cash flow (DCF);
(b) and a cost decrease per emitter for capturing CO2 (including the network costs) ranging from 4% (for the cement factory the largest emitter) to 90 % (for the compressor test platform the smallest emitter) supposing a cost ratio identical to the flow rates ratio.

For the real case studied and based on the assumptions made a limit is identify (500 ktCO2/y) under which pooling the flue gases is cheaper than having a standalone capture unit. To define such a limit for a different industrial area the same type of study must be done.

Corrosion is probably the main issue for this network with corrosion rate that can reach 2.7 mm / y for carbon steel because of the flue gas nature. Then preventive solutions must be taken in this network such as stainless steel or internal coating. Different types of polymers coatings are tested in a mixture equivalent to the flue gases and up to 60 degrees of Celsius in one-side immersion bath and autoclave. A similar protocol is applied for all the tests and the structures and the adhesion of the coatings before and after the autoclave tests are investigated using microscope and pull-off test to stay close to standards. One coating shows good corrosion performance in the test conditions, which has high corrosion resistance and no blisters are visible after the autoclave tests. It can be a candidate for application in such an environment. The thermal expansion effect on the coating deserves further investigation. Two software packages, OLI and NORSOK, are benchmarked between them and experiments. OLI is based on the complete flow composition given for each condition, including oxygen content and its model depends on the liquid film velocity. NORSOK is an empirical model based on flow loop experiments at temperatures between 5 degrees of Celsius and 160 degrees of Celsius. Its calculation is based on the mean flow velocity. There are minor deviations in the Ph prediction and larger ones in the corrosion rate. The effect of the liquid velocity is to increase the corrosion rate as some corroded particles are removed from the wall increasing the contact between sound wall and corrosive flow. Then when OLI simulations are made taking into account the liquid film velocity estimated using OLGA (multiphase simulation software) the difference between NORSOK and OLI predictions are reduced. It is very complex to make some comparison between the experiments done and the simulation as the mixture is not flowing in the experiments while it is taken into account in the software.

1.3.2 Medium pressure: CO2 from capture to Hubs

The purpose of the collecting network is to link the pooling centres where CO2 has been captured to a common hub where the export of CO2 in itself will begin either by pipeline or by ship. The overall CO2 collecting network on Le Havre and Port Jérôme areas is around 40 km long. Several options for CO2 transport up to the hub were studied:

(1) For the ship hub: CO2 must be liquefied up to the conditions (6.5 bar, - 50.3 degrees of Celsius) before loading. In order to do so three options are studied:
(a) CO2 is liquefied at the pooling centres level and then transported in the liquid form by pipelines up to the hub. Those pipelines will have to be insulated,
(b) CO2 is sent from the pooling centres to a common liquefaction unit at the hub level: (i) in a gaseous form at a low pressure,
(ii) or in a gaseous form at a higher pressure.

(2) For the pipeline hub: in order to reach the dense phase and the pressure set for pipeline transport (150 bar) there are two options:
(a) CO2 is compressed up to the dense phase at the pooling centres level and reach the hub at a pressure above 80 bar. At the hub level, pumping only is required to increase the pressure up to 150 bar before the long distance transport.
(b) CO2 is sent to the hub in a gaseous form at low pressure and is transformed in dense phase at the hub level.

In order to identify possible weak points in the network's integrity some failure scenarios are simulated on the gas transport at low and high pressure. In all the tests done including dehydration unit failure, shut down / restart of sources, leaks, emergency shut down (ESD) the following can present a challenge:

- Large leaks in dense phase network can induce very low temperatures, potentially leading to cracking of the pipe.
- ESD in gas phase scenario can lead to low temperatures in some parts of the pipe. If repeated, thermal fatigue could cause system failure.

Overall, operation of the CO2 pooling networks does not present any specific challenge that would make it impractical. Even in the case of failure of several dehydration unit components, the network can still be shut down in emergency without creating further damage other than potential corrosion issues due to the presence of water.

A Bayesian net (BN) risk assessment tool is developed to simulate loss-of-containment scenarios in the network. This model produces probabilistically weighted estimates of fatality due to CO2 exposure for a range of pressures, pipe configurations and distances from the leak source and atmospheric stability conditions. Extensive testing of different implementations in the BN software indicated that heavy mathematical equations involved in estimating plume spreading and CO2 concentrations should be pre-processed using more suitable technical software (i.e. Mathcad, Mathematica, Matlab). Pre-processing results are then imported in the BN model. Keeping the working pressure under 21 bar in the network simplifies the physics as only single phase has to be treated.

The safety risk analysis of the CO2 collecting network results in identification of significant risks at a few key locations along the collecting network. The impression of this analysis is that although leak frequencies can be expected to be small and similar to historical data in the natural gas industry, risks are none-the-less significant due to the relatively dangerous CO2 concentrations that can spread from a leak during calm atmospheric conditions. Dangerous concentrations from the CO2 collecting network are estimated to reach distances in the order of 200 - 300 meters. As there is currently no engineered means of diluting released plumes of CO2, the only options for mitigating risks after a release are maintaining safe distances from the pipelines and minimising the total mass of the release through the installation and use of “blocking” or isolation valves along the pipelines.

For our real application case (for other configurations, other distances and mapping, conclusions should be re-evaluated), the economic benchmark of the CO2 collecting network options has shown that:

- Up to the ship hub, transporting CO2 in the liquid phase is the most cost effective option. Not only does it decrease the investment (smaller pipeline) but it also has lower operating costs than in the case of gaseous transport (higher pressure drop for the latter). Moreover the collecting network size is reduced, which can play an important role on industrial sites with limited access to available space. Conversely, the space required at each pooling centre increases (individual liquefaction units). Moreover, pooling centres should have an access to cooling water, and the ammonia treatment could introduce site-specific restrictions.
- Up to the pipeline hub, transporting CO2 in dense phase is undoubtedly the most cost effective option for Le Havre case. The costs are 30 % lower in this case than in the gaseous one. Again, at the pooling centres level, there should be enough space to have a complete set of 4 stages of compressors and one pumping station.
- A common hub could be a good idea if we want to combine pipeline and ship as export systems. The economic assessment shows that the liquefaction using expansion is just 18% more expensive than the 'classic' ammonia liquefaction system (using liquid transport from pooling centre to the ship hub). This option offers a lot of flexibility. At the hub level, it is possible to envisage a pumping station to connect to the export system by pipeline and an expansion one to feed the storage tank that will supply the CO2 ships.

1.3.3 High pressure: export system to final storage (pipe, ship)

The last part of the network is focused on the CO2 transport up to its final storage. Different possibilities exist combining onshore and offshore pipes and ships. In this part of the network the quality of the mixture will have a strong impact on the network lifetime. In CCS operations, the captured CO2 stream from industrial installations is not a pure CO2: it contains some associated compounds, such as N2, O2, Ar, SOx, NOx. This mixture of gases may have significantly different thermo-physical properties as compared to a pure carbon dioxide that may impact the transportation and compression. For a global account of this impact and for a precise specification of maximal amounts of associated compounds that can be tolerated in CO2 flues, further investigations are strongly required. The present study focalises on mixtures less presented in the literature. Using Monte Carlo and Molecular Dynamics to simulate a pseudo-experimental database, and cubic equation of state to predict liquid-phase densities, CO2+NO and CO2+NO2 are simulated for temperatures ranging from (-20, 20 degrees of Celsius) and pressures (7.5 MPa - 12 MPa). Results are validated against experimental data from the literature. A special attention is paid to the degree of conversion in the NO-N2O2 equilibrium as this compound can exist as a mixture of monomers (nitrogen monoxide (NO)) and dimers (dinitrogen dioxide (N2O2)) under certain pressure and temperature conditions and we prefer to have only the monomers for the simulations. The thermal conductivities and densities of CO2+N2, CO2+O2 and CO2+H2S mixtures is also studied at 130 and 150 bar at two temperatures of 15 and 20 degrees of Celsius by non-equilibrium molecular-dynamic simulations (NEMD) and compared to results from selected EOS. The comparison for CO2+O2 and CO2+N2 showed that all models predicted the density quite well compared to the NEMD. There is a great discrepancy in the models answer regarding the CO2+H2S density. Experimental data are necessary to identify which model represents the best the reality. Regarding the thermal conductivity the pseudo-experimental data are compared with an extended corresponding state method (TRAPP), results obtained using the software REFPROP and models by Chung and Baroncini et. al . The TRAPP method shows different tendencies regarding the other models and data. Once again, even if Baroncini et al. model is very close to the NEMD data at 20 degrees of Celsius more experimental data are need to conclude on the results obtained.

A state-of-the-art review shows that as far as the historical statistics of loss-of-containment events are available, CO2 pipelines present better safety records than natural gas pipelines. Whether this can be reliably extrapolated into the future remains to be seen since CO2 pipeline networks are far less widespread compared to natural gas pipeline grids. With respect to CO2 pipelines, the following items are considered most influential for the quantitative risk analysis (QRA) outcome: phase controlled transport, failure scenarios, crater formation, CO2 probit function and modelling of dense gas CO2 outflow and dispersion.

The complexity of the physics and the first experience with the BN model leads to the conclusion that more traditional, standardised modelling approaches would be needed to analyse the export pipeline and supplemented with a fit-for-purpose computational fluid dynamics (CFD) model of a specific 'worst-case scenario'. Then CFD calculations are performed for worst case pipeline rupture release scenarios since the most commonly used simpler models for dispersion calculations do not account for terrain effects. In addition, these simpler models are not very accurate for large amounts of heavy gas dispersion where the interaction with the atmospheric boundary layer and gravitational effects become dominant. Usually, the mixing is modelled via an air entrainment factor at the upper surface of the moving cloud. With CFD, a detailed calculation of the turbulence production and destruction due to surface topology, gravitational effects and the interaction with the density gradients within in the heavy gas is performed. For both cases, low and high turbulence environment the agreement between the experimental data and the CFD solutions is good with respect to maximum concentrations and temporal behaviour. Hypothetical release scenarios are simulated that, confirm terrain topology role is important especially when the wind speed is low. The German rule VDI 3783, part 2 can be applied for maximum effect distances without taking the surface topology into account since it relates the released volume or mass flow to a wind speed which is higher than the critical wind speed. Local effect distances of up to 2 km and local characteristic times of more than 1000s can be expected from the investigated full bore rupture scenario.

A simplify model of crack propagation producing the result of complex physics that determine the final, arrested length of a pipeline crack is developed under a set of system approximations:

(a) elastic steel for the pipeline;
(b) CO2 is a perfect gas with no thermodynamic phase changes.

This model has been validated against relevant and publicly available experimental results. The smoothed particle hydrodynamics (SPH) method in the FE code Abaqus is used to analyse more precisely the mechanical phenomenon (propagation of a crack in a pressurised pipe, ejection of the fluid part, temperature and pressure change in the fluid due to the decompression). Results from this numerical model are in good agreement with the analytical model. Thanks to this numerical simulation, it will be possible, in the future, to analyse more complex geometry.

To benchmark the options to transport CO2 from Le Havre to Rotterdam a base case was defined: 13.1 Mt / y of CO2 captured from Le Havre area and transported from 2020. Designs are performed for the peak flowrate corresponding to 15.45 MtCO2 / y.

For this scenario and considering the assumptions taken for COCATE project (DR = 8 %, project lifetime = 30 years), the cost-optimised option from Le Havre to Rotterdam is the onshore pipeline (10 % cheaper than ship when only considering export system, 4 % cheaper when considering the whole chain from collection to export systems).

Ship and onshore pipeline costs were compared to what was found by ZEP for CO2 transport and it seems that COCATE costs are slightly higher than ZEP ones (essentially due to more conservative assumptions in COCATE estimate) but still in the range of those costs. From sensitivity analyses the parameters influencing the most the costs are the energy and fuel prices, and then comes the discount rate and finally the project lifetime. As pipeline is much more CAPEX intensive than ship, it is much more sensitive to discount rate and lifetime compared to ship. However, onshore pipeline is almost always the best choice under the assumptions of the project to transport CO2 from Le Havre to Rotterdam for the base case scenario.

An economic model that could be used both by governmental institutions and by a group of industrials implicated in collecting and transporting their CO2 to storage sites is developed to match the capacity left in each storage site with the CO2 transported flow rates, in order to decide how, when and where to invest in a transport facilities. The model defines in this way an optimised transport network system, involving onshore and/or offshore pipelines and/or shipping, forcing some of these transport systems to go through or to avoid nodes (e.g. harbours) with the only objective of minimising the overall costs of CO2 transport and storage, and to determine the most economical CO2 network deployment over the long term. Different scenarios are simulated, the conclusion is that the model handles correctly the user's restrictions in terms of economic results and compliance with. The results are consistent with the accuracy given by the costs functions entered by the user. Between Le Havre and Rotterdam, for 13.1MtCO2 / y transported, onshore pipeline was the cheapest solution, which is in line with other results from COCATE.

The model robustness is confirmed by a case handling additional storage sites and additional emitters.

1.3.4 Strategy of deployment Preliminary strategy for CCS deployment in Le Havre
Based on the technical, risk and economic studies presented above, a preliminary strategy of deployment is proposed for the Le Havre, Port Jérôme case.

It is assumed that there will not be enough incentives from 2020 for all the actors involved in this project to decide to launch a one-step strategy (pipeline or ship). Indeed it would require a big upfront investment as early as 2015 in order to be operational by 2020 that industrials are probably not ready to make. Consequently, a two-step deployment scenario was considered.

First step: 2020
The first industry to develop CCS on all its emissions will probably be the coal power plant and this should not happen before 2020. Indeed in 2020, the power plant will have to pay for 100 % of its CO2 emissions (ETS) and the cost of CCS chain will have already dropped. Thus the economic incentives to develop CCS on the power plant will be more interesting at this time. To those emissions, the 'easy to capture' CO2 (concentrated sources) will be added (Process Gas from the ammonia and urea production plant + process CO2 from the refineries' SMR ).

These emissions would represent 4.3 MtCO2/y that should be transported from 2020 to 2026 included. The peak flowrate used in the design corresponds to an annual volume of CO2 of 6.55 MtCO2 / y.

Second step: 2027
In 2027 all industries will have to pay for their CO2 emissions and the cost of the CCS chain should have decreased even more. The CCS chain should be implemented on the rest of the industries considered in this study.

Based on this two-step scenario, different transport implementation options are investigated (overdesign from 2020, step by step implementation for ship and pipeline, mixed export systems). For each of those strategies, the optimised designs are found for both the export system and the collecting network. In the end, we are able to compare the implementation strategies between each other through CO2 transport threshold costs and cumulated DCF.

The one-step deployment strategy (13.1 Mt CO2 / y starting 2020) seems to be the best option in terms of CO2 transport threshold costs. For the two step deployment scenario, even though there are small differences between the studied transport options in terms of transport threshold costs: around 21.5 EUR / t +/- 6 %, the best option, in terms of cumulated DCFs, is to invest in a two-step deployment strategy with a two-step investment for ships (option no 5).

All in all, it is difficult to point out one strategy as the best for the COCATE implementation scenario through these basic costs calculations. Using macro-economic analysis such as business models and real option analyses should help orienting the choice for the best implementation strategy. But it would require more precise and more detailed studies. However results are encouraging enough that they can be shared and useful for other industrial clusters.

A first step is therefore to identify cluster in Europe and a second step to propose a generic recommendations for the business model that could supplement the results obtained previously. Other cluster in Europe and South Africa
In order to identify such areas, the following limitations, revealed by the studies performed in COCATE, are considered:
(a) the number of industrial CO2 emitters in the same geographic area;
(b) the total amount of CO2 emitted that can be sufficient to justify the viability of such a network;
(c) the variety in the types and sizes of industrial sites (power plant, cement, petrochemicals, incinerators) to promote a steady and continuous capture.

The goal of this study is not to identify all the possible clusters in Europe, but to put in light the fact that there is a great number of small emitters located on restricted areas that could be interested in sharing the capture and conditioning cost together or with a bigger emitter. For this research a methodology is defined and validated on the French case and then applied to Europe and South Africa.

This methodology is tested for France in order to check if the cluster of Le Havre appears within the database results.

Results showed that the Upper Normandy region, including Le Havre, appears fifth. The emissions generated by the top five regions represent 60 % of the total amount for CO2 emitted in France.

For these five regions emitting high levels of CO2, a visual research on possible cluster is done in order to identify the number of sources located on areas smaller than 400 km2. Five clusters are identified:

- Dunkerque, Region Nord Pas de Calais,
- Marseille-Fos, Region Provence Alpes Côte d'Azur,
- South Lyon - Region Rhône Alpes,
- Nancy, Region Lorraine,
- Le Havre, Region Upper Normandy.

As Le Havre appears as a potential cluster, the methodology can be used to identify potential clusters where flue gases could be pooled and CO2 captured in shared units.

Once the methodology is validated, it is applied to other countries than France in Europe. In 2010, 2 Gt of CO2 emitted in Europe are reported in the database. In order to reduce the study to countries with the greatest potential for clustering, only the countries with the largest CO2 emissions are selected. The search is then narrowed in order to identify the region with the greatest CO2 emissions.

8 European countries come out of the study, they are listed below with the regions within the country were the greatest clustering potential is identify:

(1) Germany:
- Düsseldorf area , most potential around Duisburg,
- Köln area,
- Brandenburg, Münster, Hamburg area, most potential around the port of Hamburg.
(2) United Kingdom
- Derbyshire and Nottinghamshire area, 2 clusters possible,
- North Yorkshire area,
- Eastern Scotland (Edinburg and Fife estuary) area,
- East Yorkshire area, most potential around the port of Kingston Upon Hull,
- Cheshire area, around the port of Liverpool,
- West Wales and the Valleys area, around Milford Haven,
- Tees Valley and Durham area, potential cluster around the Redcar, Middlesbrough port area,
- Kent (Thames Estuary) area.
(3) Poland
- Silesia,
- Opole.
(3) Italy
- Apulia, area around the Port of Taranto,
- Sicily, area around the Port of Syracuse,
- Lombardy, even though a specific cluster is not clearly identified,
- Vento
(4) Spain
- Andalusia, 2 clusters came out: the Huelva Port are and the Port of Gibraltar,
- Asturias,
- Catalonia, around the Port of Tarragona,
- Galicia, around the Port of La Coruña,
- Basque Country, around the Port of Bilbao.
(5) The Netherlands: Port of Rotterdam
(6) Belgium:
- Port of Antwerp.
In South Africa the Durban area including 60 sources emitting less than 500 ktCO2 / y and 10 sources emitting between 500 and 1 500 ktCO2 / y could find an interest in pooling their flue gases to reduce the capture costs.

This shows that many areas in Europe and South Africa could be fitted to host mutualised CCS projects; a generic basic business model as well as decision tree is therefore developed. Recommended workflow

Since it is shown that other areas in France as well as elsewhere in Europe and South Africa could be fitted to host a mutualised CCS chain, it is decided, based on COCATE experience and main conclusions, to develop a workflow.

This workflow is a general framework to organize the development of a CCS project which includes capture installations, pooling and conditioning infrastructure of fumes, CO2, and solvent, hub(s) facilities for exporting CO2 to storage site(s). Considerations on capture and storage have been included since they are fundamental items of the decision making process.

The results are presented through two useful tools. First, a checklist necessary to analyse each part of the CCS project chain has been created, second, a workflow organising the different steps of an efficient development strategy for shared CO2 transport network. Below is presented the general outline of the workflow followed by the detailed steps:
The first sequence of action concerns the emission analysis and the capture pooling and clustering. This should be performed in parallel with storage capacity analysis since it might impact the overall development program.

The second sequence of action concerns the CO2 hub(s) and collecting network(s) and the export system(s). The CO2 hub and collecting networks analyses should be performed simultaneously with the export system design since one depends on the other. Moreover, one should also mention that storage development strategy is strongly needed at this stage to help shape a network which will be able to cope with storage parameters. Business models of the CCS chain
More than a business model, this section presents the main concerns to be addressed when analysing the business aspects of alternative physical CCS configurations in an industrial zone.

The main premises when defining the business environment for developing a CO2 collection system in an industrial zone are listed below:

(a) the conceptual design of the hardware of the CCS chain;
(b) the CCS actor framework, with the roles per actor:
(i) large emitters;
(ii) small emitters;
(iii) collection system operator;
(iv) transport system operator;
(v) storage operator;
(c) the pertinent contractual relationships between the actors;
(d) the regulatory regime, if applicable;
(e) the business drivers and tariffs.

These are the main categories, actors, to consider when developing a CCS chain on a specific area. However one should keep in mind that this type of collaboration is new and nothing similar has been done yet.
Therefore this list is probably not exhaustive.

1.3.5 Conclusion and recommendations

- Low pressure: flue gases (and alternative)
Obtaining the information on future emissions is very difficult as it involves strategic data that emitters are not willing to share between each other, the site longevity is unknown: close down, relocation, newcomers are difficult to identify.

Difficulties encountered in making the routing and design of the flue gas collecting network are:

- Territories evolve: it would be interesting to associate local actors/authorities to the project in order to map the current and future constraints more accurately.
- Location of the emitters: in the case a small emitter is located in the middle of a populated zone it cannot be possible to include it in the pooling strategy.
- Designing networks for the peak flowrates not best choice it could be more interesting to design the network at an optimised flowrate (if the emitter gives the flow pattern) to limit the overdesign and to capture the greatest volume of CO2 (an optimum should be found between the overdesign cost and the cost of emitting CO2).
- Check for day / night profiles for flow rate variations.
- If flue gas flowrate is above 105 m3 / h it becomes interesting to consider the common regeneration concept. If there is no absorption column in the vicinity (maximum 150 m) of this type of source, add an absorption column.

Recommendations for future studies:
For the source description: it is important to take time to present the project to each industry, allotting enough time for the questionnaire completion, giving incentives to the industries and having two contacts within each company: a technical one and one aware of the company's strategic decisions for the future. If capture technology is defined then suggest two types of flue gas collection networks: one grouping together the flue gases that can go directly to the capture unit (no additional treatment) and another one going through a pre-treatment centre before capture. If a post-combustion technology with solvent is chosen for capture, study other options for pooling or try to combine options:
(a) common regeneration centre;
(b) common conditioning centre;
(c) common 'energy' centre.

To identify when pooling should be considered it is necessary to study each industrial case as this depends on:
(a) the volume of flue gases;
(b) the CO2 concentration;
(c) the volume of CO2 captured.

Pipelines construction material should be chosen with care to limit corrosion during flue gas transport or internal coating should be applied in common steel pipes.

- Medium pressure: CO2 from capture to Hubs.
- A special care about the safety risk considerations has to be paid as: loss of containment of CO2 can be hazardous at up to several hundred meters from the leak source in some plausible scenarios:
(a) this will limit where the collecting network can be routed in relation to third parties;
(b) additional effort on consequence modeling is justified in some situations with complex terrain and buildings, and when fundamental failure modes and initiators are relatively uncertain;
(c) CFD in addition to analytical modelling; (d) BN safety risk model in addition to fault / event trees and pre-defined leak scenarios;
(e) enhanced safety features should be considered early;
- Therefore the routing choice is an iterative process which should include all potential mitigating design choices that can reduce minimum safe distances to the collecting network. The design process may therefore require more time and effort to complete. Options for enhanced safety features are:
(a) re-routing to increase distances to third parties;
(b) thicker pipe walls than standard to withstand impacts and provide increased corrosion safety margins;
(c) buried pipe deep enough to minimise probability of impacts from machinery;
(d) pipe elevated up to 10m above ground to increase air dilution in loss-of-containment;
(e) close placement of emergency shut-in valves with automatic actuators;
(f) permanently installed blowers or stand-by helicopters to disperse large CO2 releases when atmospheric conditions are stabile;
- Detailed work on CO2-specific issues in safety risk analysis has shown the following areas recommended for further work:
(a) potential to improve analytical dispersion models to reduce gap with CO2 release experiment observations;
(b) additional laboratory studies to achieve consensus on Probit parameters for CO2;
(c) reduce pipe failure frequency estimates by continued focus on recommended pipe material choice and wall thickness standards for CO2;
(d) running ductile fracture probability must be robustly minimised in the design phase;
- High pressure: export system to final storage (pipe, ship).
Regarding the impact of the impurities on the flow behaviour, new experimental data are needed for:
- H2S as function of temperature and pressure (not only at saturated conditions) to study binary systems,
- CO2+NOX mixtures,
- CO2+SO2 mixture. The presence of water should be also performed with the corrosion problem.

It seems also important to do similar studies on viscosity as for thermal conductivity and diffusion coefficients. The viscosity is an important property for both transport and for design of process equipment like for instance heat exchangers.

Regarding the cost sharing and tariff issues, the blocking points identified are:

(a) decisions by the industrials on the criteria to be used to determine the cost allocation is of the essence to determine convenient cost allocation rule;
(b) from a tariff issue and policy maker standpoint, allowing price discrimination is the best way to ensure participation of all emitters to infrastructure;
(c) pipeline utilisation on the long term must be forecasted to build appropriate rules this should meet design concerns;
(d) CCS project can have a negative NPV which is a difficulty for the real option modelling.

Reducing technical uncertainties would help better value the different options within actual and future political framework.

Potential impact:

As coordinator, IFPEN has developed a web site to disseminate COCATE activities during the project lifetime to the public and one to maintained confidential exchanges between the partners. Both web sites will be maintained after the end of the project. Thermodynamics work done in COCATE has reinforced the internal knowledge on the impurities impact on phase diagram and transport properties. This work has also put in light some missing experimental data which will be useful for future projects. The economic optimisation model developed can be helpful for governmental institutions and for groups of industrials implicated in collecting and transporting their CO2 to storage sites in order to decide how, when and where to invest in a transport facilities. The work done on the crack propagation and the fluid/structure interaction will support future project on CO2 or gas transport and storage.

Accoat A/S expect to benefit from the main dissemination activities and exploitation of results by its use for emphasising company's green profile and its environmental policy as a provider of surface solutions that reduce the strain on the environment and the energy use. Activities and results of this work are being used to underline company's approach towards clean-tech applications and markets, as well as for increasing environmental consciousness among Accoat's suppliers and customers together with the financial sector.

The South African Centre for Carbon Capture and Storage SACCCS (a division of the South African National Energy Development institute SANEDI) is the prime mover for carbon capture and storage in South Africa. The main thrust of CCS work in South Africa is a Pilot CO2 Storage Project planned for first injection during 2017. As such, the issue of transport has been delegated to a lower priority, the outputs of the COCATE project being relied upon for information in this field. SANEDI's participation in COCATE was to bring a developing country's perspective and to facilitate capacity building in the country. To this end, an information dissemination workshop was held in Johannesburg 7 - 8 November 2012 followed by a field trip by the international visitors to the Sasol plant at Secunda, the largest point source of high concentration CO2 in the world. During the first day of the workshop, general transport of CO2 (pipelines, ships, tankers) was addressed with most of the presenters being from the COCATE project. The second day focused on the COCATE project results. The topics of the Workshop were new to the 60 participants, thus building capacity as enablers for further CCS work in South Africa.

In particular, the aggregation of small emitters in the Durban area was undertaken as a COCATE Project case study. The Durban area was chosen as it has a large number of small and medium size emitters and is close to prospective storage sites as well as having the best information database. Once storage capacity in the area is proven, this case study may be used as a basis for implementation.

SINTEF Energy Research plans for exploitations of results include:

(a) the use the increased knowledge about the technical and economic evaluation of CO2 conditioning and transport for assessing CCS value chains in further research works;
(b) the use of the methodology developed to compare modelling of new gas mixtures properties (for example CO2 + impurities);
(c) the use of the series of reports and works published as reference documents in future works, applications for new projects, and presentations.

LHD expect that the dissemination of some of the results to the industrial partners will fast-forward:

(a) in the short term the CO2 capture, transport and storage research initiatives;
(b) in the longer term, the construction of a CCS demonstrator in Le Havre.

Working on the COCATEs case study, TNO was involved in all of the projects work packages. The opportunity was there to show the variety of in house expertise's the Dutch organisation for applied research has to offer, all along the whole carbon capture transport and storage (CCTS) chain. In doing so, TNO has strived for much knowledge sharing, as is reflected in several (recently submitted and accepted) peer reviewed journal (scientific) articles. This knowledge development is not considered only from a theoretical angle. TNO certainly wishes to further implement this knowledge from theory into practice, and will likely assist industrial partners in CCTS (pilots, demo) projects.

Major efforts have been achieved on two focus point: corrosion issues, and safety. Amongst other topics (CO2-capture, subsurface - geology etc.), and after COCATEs closure, TNO research will further proceed on 'tailor-made' dense gas dispersion modelling (CFD, including terrain effects - in a built environment) in order to solve some of the remaining CCTS risk issues left.

List of websites: