J.S. Gudmundsson* and A. Børrehaug**
* Department of Petroleum Engineering and Applied Geophysics
Norwegian University of Science and Technolog
7034 Trondheim, Norway
** Aker Engineering
0250 Oslo, Norway
The frozen hydrate method is considered suitable for large scale transport of natural gas over long distances. The results from a recently completed feasibility study confirm the physical basis of a natural gas hydrate chain (production plant, hydrate carriers, regasification plant). An estimate of the capital cost of a natural gas hydrate chain for the transport of 400 MMscf/d over 3500 nautical miles was found to be 1813 million US dollars. The capital cost of the frozen hydrate chain was 24 percent lower than the capital cost of an equivalent LNG chain.
Frozen hydrate can be used to transport large volumes of natural gas over long distances. The frozen hydrate concept is based on the discovery that natural gas hydrate remains stable at atmospheric pressure when stored under near-adiabatic conditions at temperatures below the freezing point of water (1). A natural gas hydrate (NGH) transport chain comprises hydrate production, marine transport and regasification elements. Laboratory work and feasibility studies indicate that our NGH chain has the potential of reducing the delivered cost of natural gas considerably compared to established LNG technology. The purpose of the present paper is to present the overall results of work at Aker Engineering in Oslo and the Norwegian University of Science and Technology (NTNU) in Trondheim, on the use of frozen hydrate to transport natural gas in large volumes over long distances. The paper draws on results from a feasibility study of a NGH chain for the transport of 400 MMscf/d of natural gas over 3,500 nautical miles, assuming European conditions.
Our laboratory work on frozen hydrates dates back to 1990. Experimental results carried out 1990-1991 were published in the open literature in 1994 (1). The work showed that natural gas hydrate did not decompose when stored at temperatures in the range -15 to -5 deg. C in an ordinary freezer at atmospheric pressure. Since then the laboratory work has been extended, with the same overall conclusions. Co-operation between Aker Engineering and NTNU was initiated in 1992. In early-1993 an internal study by Aker Engineering (Transport of Natural Gas in Hydrate Form, A. Børrehaug) identified cost saving to warrant further study. In 1994 we carried out a joint design and cost study for a NGH chain and compared the costs to a LNG chain (Hydrate Chain Compared to LNG Chain, F. Hveding). This work was the basis for a and paper presentation in 1995 (2) The overall conclusion of our 1994 work and 1995 publication was that the capital cost of the NGH chain was 25 percent less that of the LNG chain. The NGH chain costs were based on Aker Engineering cost estimates while the LNG costs (adjusted to mid-1994) were based on DiNapoli (3). The case examined was transport of 400 MMscf/d over a distance of 3,500 nautical miles. We selected these conditions because of planned developments in northern Norway. Subsequent work at Aker Engineering and NTNU has used the same assumptions.
Our work on frozen hydrate has been driven by the growing need for natural gas in Europe and Asia Pacific (4, 5). Natural gas needs to be transported into these regions over increasing distance; for example, the Barents Sea to continental Europe and Western Australia to Japan.
The NGH project at Aker Engineering and NTNU is based on two main premises. First, that frozen hydrates are stable enough to be stored in large volume and transported over long distance. Second, that a NGH chain has the potential of delivering natural gas to market at significantly lower cost than a LNG chain. The first premise is our foundation, while the second premise drives our work. In addition to being stable, the frozen hydrate must contain as much natural gas as possible.
In our laboratory at NTNU we have produced hydrate in a high-pressure reactor operated at 50-70 bara and 2-10 deg. C. Typically, our 600 cm3 reactor was charged with 100 cm3 liquid water and 500 cm3 natural gas mixture, containing 92 mol % methane, 5 mol % ethane and 3 mol % propane (1). Hydrate formation started when a magnetic stirrer was activated and operated at about 500 rpm. The formation of hydrate was monitored by measuring the fall in pressure with time (reactor in constant temperature bath). After about one hour the stirrer was turned off and the reactor turned up-side-down (in the constant temperature bath) to separate the liquid water and solid hydrate. Various separation methods have been tested, both gravity separation and filtration. We have observed that the remaining hydrate contains free water.
After the separation the reactor was moved to a deep freezer operated at about -18 deg. C. The reactor was left in the freezer over night and then the pressure was quickly lowered to atmospheric pressure. The hydrate was removed and immediately placed in an ordinary (air filled) freezer. The stability of this frozen hydrate has been monitored with time.
Based on our experience in the laboratory, and general understanding of hydrate properties, we have selected the following design values for use in our NGH hydrate project: (1) gas content of pure hydrate 150 Sm3/m3, (2) free (additional) water as ice 10 wt. % and (3) bulk porosity of hydrate/ice mixture 25 vol. %. Overall, the frozen hydrate/ice mixtures will contain about 100 Sm3 natural gas per m3 of bulk volume. This value is much lower than the 170-180 Sm3/m3 gas content value reported in the literature for pure hydrate. The heat of formation and decomposition of produced hydrate was assumed 410 kJ/kg. The specific heat of natural gas hydrate was assumed 2 kJ/kg.K and its density 948 kg/m3. The 150 Sm3/m3 value is based on less than theoretical degree of filling of the hydrate structure. The 10 wt. % free water is based on less-than complete separation of solid hydrate from liquid water. The 25 vol. % porosity is due to the packing density of frozen hydrate as particles in bulk.
The NGH chain consists of three main parts: the production, the marine transportation and the regasification. The production part is assumed to be located on land with loading facilities for large hydrate carriers. The transportation part is by bulk carriers specially designed for frozen hydrate. The regasification part of the frozen hydrate takes place at a receiving terminal located close to markets for natural gas.
The design of each of the three parts of the NGH chain has not been fixed in the on-going work of Aker Engineering and NTNU, and it will continue to evolve. We will present a version here called the "modified case" in the feasibility study completed by Aker Engineering in late-1995 (Natural Gas Hydrates for Large Scale, Long Distance Gas Transportation).
The NGH chain was designed for 400 MMscf/d (about 11.3 million Sm3/d) of natural gas. The production plant was assumed to have 4 parallel trains, each 100 MMscf/d in capacity. A schematic overview of the hydrate production process is shown in Figure 1. Process flow diagrams with heat and mass balances were prepared for the production plant. The separated natural gas arrived to the hydrate reactor at 65 bara and 10 deg. C with the following composition: C1 94.0 mol %, C2 4.6 mol %, C3 1.4 mol %. The condensate recovered from pretreatment was not considered further in the hydrate production plant. Before the natural gas is passed to the reactor system, the gas is used (not shown in Figure 1) in the drying and freezing operations down-stream of the hydrate/water separation system.'
In the NGH production process, it was assumed that fresh water at 2 deg. C is pumped into a Continuous Stirred Tank Reactor system operated at about 10 deg. C. The natural gas arrives from the drying and freezing system (see below) and is injected into the system (reactors in series) and hydrate forms under conditions of active mixing. The liquid water comes from an ammonia-based cooling system where 9,500 tonnes/hour of water are cooled, corresponding to 84 MW thermal. Seawater was assumed as heat sink and the ammonia loop compressor duty was estimated 23 MW. Note that these numbers are per train and that power duties and related costs depend on the seawater temperature assumed (5 deg. C in present work). From the reactor system the hydrate/water slurry in each train flows to three belt-filter separators.
The hydrate leaving the separators is assumed to contain about 12 wt. % free water. The wet hydrate is passed through a rotating dryer where counter-current dry natural gas forms hydrate, reducing the free water content to about 10 wt. %. The hydrate is then passed through a screw-type conveyor where the hydrate is refrigerated to about -15 deg. C.
The mechanical design of the dryer and freezer have not been specified. In our work, however, we have made the necessary heat and mass balances, and then estimated roughly the costs based on similar equipment in the pulp and paper industry. The counter-current natural gas flow requires compression and water removal with the necessary cooling and separation equipment. Both the power requirements and equipment costs associated with this part of the production process have been included.
The -15 deg. C hydrate is conveyed to a pressure reduction tank. There are three such batch- operated tanks for each train. At any one time one tank is filling, one tank is reducing pressure and one tank is emptying. The pressure is reduced from about 50 bara to atmospheric pressure. The frozen hydrate is conveyed from the emptying tank by conveyors to a hydrate carrier. The hydrate is at atmospheric pressure, -15 deg. C and surrounded by natural gas. Equipment and conveyors are enclosed and insulated to maintain the required temperature.
Outline specification of a hydrate carrier has been prepared by Shipping Research Services in Oslo. The carrier designed is 357 m long, 60 m wide and with 20 m draught. The net loading capacity is 322,000 tonnes and the maximum cargo volume 460,000 m3. The design service speed is 15 knots in transit and 15.8 knots in ballast. The hydrate carrier has 12 box shaped cargo tanks and 12 ballast tanks (140,000 m3) and has double bottom. The cargo tanks have limited insulation, primarily to prevent ice formation on the outside. The general design specifications are for a bulk/oil carrier.
The transportation part of the NGH chain involves much larger volumes than in the LNG chain. One cubic meter of LNG is equivalent to about 600 Sm3 of methane. If the gas content of natural gas hydrate is 150 Sm3/m3 the NGH volume is 4-times that of LNG. If the bulk gas content of natural gas hydrate is 100 Sm3/m3, as discussed above and assumed in our feasibility study, the NGH volume is 6-times that of LNG.
We have assumed that NGH carriers need to be much larger than LNG carriers which are typically 125,000 to 135,000 m3 LNG. NGH carriers are likely to be 3 times larger than LNG carriers and are therefore more like Very Large Crude Carriers. Two NGH carriers would then carry the same volume of gas as one LNG carrier. We note however that NGH carrier design is not volume constrained but mass constrained.
The off-loading of frozen hydrate raises some challenges. A NGH carrier can simply be loaded from the top by conveying of frozen hydrate in enclosed tunnels from the storage tanks. The bulk load will be frozen hydrate with natural gas filling the pore space and ullage space. During the voyage, heat will flow into the cargo through the tank walls and cause local decomposition of the frozen hydrate into gas and ice. The liberated gas will be used for engine fuel, and will quickly be reduced by the formation of an insulating ice layer against the tank wall.
Mechanical self unloaders were assumed in an early design of a NGH carrier. Each of the 12 tanks was provided with a mechanical unloader, similar to unloaders currently used for wet coal and other bulk cargo. However, we realised that it would not be cost effective to install expensive equipment in each of the 12 tanks of several NHG carriers, because each unloader would only be used for the few days of unloading.
The transportation distance assumed in our feasibility study was 3500 nautical miles. We estimated that each round trip would take almost 30 days (loading 4 days, unloading 4 days, transit journey 10 days, ballast journey 10 days, waiting 2 days). For a NGH hydrate chain for 400 MMscf/d and based on the hydrate carrier designed, we estimated that 7 carriers were needed. This compares to 3 LNG carriers needed for the same duty (2).
In our current NGH carrier design we assume that the frozen hydrate is melted in the carrier. A schematic overview of the modified case hydrate regasification process is shown in Figure 2. Water at 20 deg. C is pumped into the 12 cargo tanks of the hydrate carrier. The internals of the carrier tanks for the hydrate melting have not yet been designed. However, preliminary considerations indicate that such a system is feasible and not expensive. The hydrate melts in the tanks and the natural gas is ducted in large-diameter ducts from the carrier to compressors on land. The water is pumped from the NGH carrier to a water heating system based on heat recovery from compressors and heat pumps.
We have assumed one compression train with a capacity of 400 MMscf/d. The natural gas ducted from the hydrate carrier is compressed from atmospheric pressure to 80 bara. The total compression duty is about 100 MW, supplied from a combined gas and steam turbine system.
The 20 deg. C melting water needed amounts to about 38,000 tonnes per hour. Water at 10 de. C is heated as follows: heat pump system (compressor duty 39 MW, heating duty 203 MW), condensing steam down-stream of the steam turbine (203 MW), additional smaller steam condenser (24 MW) and heat from export compressor after-coolers (100 MW).
The natural gas is dewatered before export in a glycol dehydration unit. A water dew point of -6 deg. C at 80 bara is assumed. The regasification plant includes export gas metering and the necessary utilities.
Above we have given a general description of a NGH chain for the transport of 400 MMscf/d of natural gas over 3500 nautical miles. For the most part, the design was based on known technology, or technology that can be developed from known engineering principles and practice. A feasibility study has been carried out in sufficient detail to estimate the sizes and costs of major components in the NGH chain. Established costing procedures have been used to arrive at the total cost of the NGH chain.
Our aim has been to find out to what extent NGH technology has the potential to compete effectively with LNG technology. We decided that a comparison based on capital costs would be suitable for this purpose. A comparison based on natural gas transfer costs (for example, cost per million Btu or gas volume) would require too many site specific and project specific assumptions, including interest rates.
The capital cost (US dollars) of the NGH chain is shown in Table 1. Also shown is the capital cost of a comparable LNG chain. The cost numbers are for 400 MMscf/d over 3500 nautical miles, based on mid-1995 conditions in Europe. The NGH costs are from our design and feasibility study. The LNG costs are also from our feasibility study, but were estimated outside Aker Engineering. The LNG chain is based on 3 ships (each 250 million US dollars) and up-to-date technology for production and regasification.
We observe that the NGH chain is 24 percent lower in total capital cost than the LNG chain. Furthermore, the dominant cost reduction stems from the production part of the chain. The cost of transportation (shipping) is similar, the NGH option being 6 percent lower. The regasification part of the NGH chain is 21 percent lower than that of the LNG chain.
We consider the NGH chain presented above a conservative design that will work technically and can be costed. The feasibility study has identified what technical assumptions and equipment and components need to be considered in on-going research and development work. We estimated that about 80 percent of the NGH chain capital costs can be assigned to known equipment and technology, while 20 percent relates to technology specific to natural gas hydrate. On-going and planned work will address both the technical assumptions and design of NGH specific equipment and components.
The major technical assumptions in the NGH chain are the gas content of the natural gas hydrate, the amount of free water and porosity of frozen hydrate in bulk (150 Sm3/m3, 10 wt. % and 25 vol. %, respectively). The theoretical gas content of natural gas hydrate at the pressure and temperature planned in production process, is in the range 170 to 180 Sm3/m3. Assuming an average gas content of 170 Sm3/m3 can be reached, the overall gas content in the hydrate carrier (assuming same percentage free water and bulk porosity) will be 115 Sm3/m3. The direct consequence will be that the number of hydrate carriers can be reduced from 7 to 6, a saving of about 100 million US dollars. Therefore, the NGH chain will be 28 percent lower in capital cost than the equivalent LNG chain.
In the NGH chain design presented above, the pressure reduction in the production process is assumed to take place in 3 large tanks for each of the 4 parallel trains. This makes the pressure reduction operation an expensive batch operation. We have made a preliminary evaluation of the possibility of using techniques developed in solids handling of fertilisers for continuous pressure reduction. Technical solutions exist and are in operation but need to be modified for frozen hydrate. We expect improved ease of operation and significant reduction in capital costs.
The off-loading of a hydrate carrier is an important challenge in the NGH chain. In the present design the frozen hydrate is melted in the carrier. While this technical solution is feasible, it may not be the best. For example, it requires compression of the gas from atmospheric pressure to 80 bara. We are currently evaluating the possibility of using slurry technology based on pumping to off-load a hydrate carrier. The frozen hydrate would be ransported in dry form, but would be mixed with condensate at the receiving terminal. The tanks would be flooded with condensate (at same temperature as frozen hydrate) and then emptied using jet pumps and slurry pumps. Most of the motive power would be land-based.
The hydrate carrier tanks would be flooded in turn and the condensate recirculated. The solid hydrate and liquid condensate would be separated mechanically and the solid hydrate and remaining condensate heated and separate at 10 to 20 bara pressure. Compression power would be reduced considerably and gas/liquid separation would take place on land. As the NGH will be offloaded as a cold slurry, storage is possible prior to regasification. The use of slurry technology opens also for the possibility of offshore off-loading of a hydrate carrier. We expect slurry-based technology to offer savings in operating and capital costs.
| Chain | LNG (million US) | NGH (million US) | Difference |
| Production | 1220 (51%) | 792 (44%) | 428 (35%) |
| Carriers | 750 (32%) | 704 (39%) | 46 ( 6%) |
| Regasification | 400 (17%) | 317 (17%) | 83 (21%) |
| Total | 2370 (100%) | 1813 (100%) | 557 (24%) |
