Introduction
Safe and reliable compression of hydrogen, CO2, and ammonia is not a new venture. In the quest for more sustainable energy carriers to facilitate a transition from fossil fuels, hydrogen, ammonia, and synthetic gases have shown promising results and garnered growing political interest. However, present solutions need to scale up on several fronts to replace fossil sources and meet the sustainability targets set for 2030.
Beginning with hydrogen production, various low-carbon technologies are available, each differing in technology readiness and scalability. These technologies can deliver hydrogen in larger quantities at low to moderate pressure. For transportation, hydrogen can be compressed, liquefied, transformed into ammonia or synthetic methane, blended in existing pipelines, or embedded into liquid organic carriers.
In most cases, the hydrogen must be further compressed, with different technological challenges and different requirements (pressure and flow) for each transportation strategy. As demonstrated in the sections that follow, a single compression technology may not always be the optimal choice.
General overview of hydrogen compression technologies
Hydrogen compression is widely adopted and referenced in refinery and petrochemical processes, providing a solid foundation for applications in the emerging green hydrogen sector. The greatest technological challenges include handling high pressures, preventing gas contamination, using materials resistant to embrittlement, and addressing the expected volumetric flow increase and fluctuations in operation resulting from the market’s scale-up and intermittent renewable power production.1 Depending on flow rate and pressure ratio, various compression technologies are available.
Reciprocating and centrifugal compressors are the most widely adopted solutions for a broad range of hydrogen applications. Reciprocating compressors (RC) have been used in refineries for decades, effectively compressing high-content hydrogen blends. Compared to centrifugal compressors, reciprocating compressors can achieve pressure ratios up to 100 and can partialize flow rates while maintaining efficiency. As a volumetric machine, the reciprocating compressor is ideal for small to medium flows, while centrifugal compressors are more suitable for larger flows.
With the increasing flow rates in new hydrogen applications, centrifugal compressor technology is expected to play a key role. However, due to the low molecular weight of hydrogen, the typical pressure ratio (PR) of hydrogen centrifugal compressors does not exceed 1.3, usually proving insufficient for pipeline applications or large hydrogen production projects. To improve centrifugal compressor performance with hydrogen, an optimized impeller aerodynamic design and higher tip speed are required to achieve a larger pressure ratio. This technology, available within Baker Hughes as ‘High Pressure Ratio Centrifugal Compressor’ (HPRC), is capable of reaching a PR of up to 2 in a single body with dry hydrogen.
For the same compression duty, HPRC technology results in smaller and lighter units, reducing the number of required impellers, and, consequently, the number of bodies per compression train. This leads to improvements in plant footprint, reliability, availability, and weight, providing significant benefits in terms of project CAPEX and OPEX.
Compressors for green hydrogen production
Green hydrogen, produced through water electrolysis, needs to be compressed to 200–700 bar for transportation, storage, or process injection. Depending on hydrogen flow and pressure ratio, reciprocating compressors, centrifugal compressors, or a combination of the two (a hybrid solution) may be required.
For small to medium green hydrogen production plants (electrolyzer capacity up to 200–300 MW) delivering hydrogen at atmospheric pressure, reciprocating compressors are the best fit. Centrifugal compressors begin to have a role in larger green hydrogen production plants.
If the off-taker requires high delivery pressures combined with large flow rates, the optimal solution is to combine centrifugal and piston technologies with a ‘hybrid’ solution. In this configuration, the first part of the compression is done by centrifugal compressors, capable of larger flows, and the remaining duty by reciprocating compressors that can easily increase the pressure to the required value. A solution based only on centrifugal compressors would allow effective management of the flow rate but not the delivery pressure (multiple casings in series). On the other side, the use of reciprocating compressors would only be a good fit to manage pressure but would require a considerable number of units in parallel.
Assuming a green hydrogen production plant with an electrolyzer capacity of 2 GW and required compression capacity of 1,500 TPD (wet hydrogen, water-saturated at atmospheric pressure), the configurations with a full reciprocating compressor and the optimized hybrid solution are shown in Figure 1. Benefits of the hybrid solution in terms of CAPEX, OPEX, and footprint are evident.
The next development in hydrogen compression with centrifugal compressors will be an additional power density step up to a PR of 3 per body with dry hydrogen. This would enable a full centrifugal solution for the above case study, with additional benefits in terms of total cost of ownership and footprint.
Compressors and turboexpanders for hydrogen transportation
Currently, hydrogen transportation can be performed leveraging:
- Liquefied hydrogen
- Hydrogen pipeline
- Ammonia
Compression technology plays a primary role in all of the above processes.
Hydrogen liquefaction
While hydrogen liquefaction is a technology first experimented with in 1898, industrial processes were developed in the twentieth century.2 Liquefaction processes are characterized by one or more refrigeration cycles to decrease feed gas temperature to its liquefaction point.
The simplest way to cause the temperature drop is a Joule Thompson (JT) valve: the refrigerating gas expands, decreasing its temperature (as long as it is below the Joule Thompson inversion temperature). Compressors are needed to increase the gas pressure before the expansion.
This low-complexity process design can be easily improved: JT valves can be replaced by turboexpanders, systems that expand gas through a different thermodynamic transformation (quasi-isentropic instead of isenthalpic) with a lower final temperature, providing a clear advantage for the liquefaction process.
Moreover, it is possible to recover the energy otherwise dissipated during the gas expansion with the introduction of turboexpander-compressors or turboexpander-generators. The former recover energy to drive compressors, while the latter are connected to electric generators, transforming the expansion into electric energy, further increasing process flexibility and efficiency.
Usually, smaller plants are equipped with reciprocating compressors and turboexpander-compressors, while larger plants are expected in the future to leverage centrifugal compressors and turboexpander-generators. As an example, Table 1 shows the products selectable from the Baker Hughes portfolio for different sizes of hydrogen liquefaction processes.
Hydrogen pipeline
Existing pipeline compressors are designed for natural gas. Specific evaluations have to be done in terms of material selection and performance when dealing with hydrogen or hydrogen blends.
Two main scenarios are currently being evaluated: dedicated infrastructure for 100% hydrogen transportation, or repurposing existing infrastructure for blending in natural gas (up to 20% volume). While repurposing existing pipeline stations for blends up to 20% hydrogen is expected to require minor adjustments, pure hydrogen compression requires dedicated design.
Hydrogen pipelines are expected to operate in a pressure range of 70–100 barg. Small pipelines for local distribution from the production site to the off-taker in a nearby area will be provided with injection compression station and, at most, one boosting station. Hydrogen flow is limited, and compressor operation in the injection station is strictly related to electrolyzer production output. Reciprocating compressors are suitable for this case, as they are capable of providing large pressure ratios and capacity flexibility.
Different considerations apply to large pipelines developed across thousands of kilometers of distribution grid to connect production and off-takers in different countries. In this case, the infrastructure will include injection compression stations and several boosting stations. Large pipeline infrastructures will also be used as storage medium, capable of self-balancing fluctuations in injection points. Large flows are foreseen, positioning the centrifugal compressor as the optimum technology, as widely applied in natural gas pipelines.
In the following case study, a comparison of different centrifugal compressor solutions to optimize hydrogen pipeline boosting stations at a constant flow of 2,000 MMSCFD, inlet pressure of 60 barA, and discharge pressure 110 barA is shown.
With natural gas and a typical impeller peripheral speed of 250 m/s, a single compressor with three impellers is sufficient. As the percentage of hydrogen in the pipeline increases, the number of required impellers also increases, reaching up to 28 for the 100% hydrogen case (five compressor bodies).
Using high head impellers with a peripheral speed of 300 m/s reduces the number of compressors to three (18 impellers). With HPRC technology, it is possible to reach a peripheral speed range of 400 m/s, thereby reducing the number of impellers to nine for the 100% hydrogen case. In this scenario, a single compressor body is enough, as in a typical natural gas pipeline.
Ammonia
Ammonia is industrially produced using the Haber-Bosch process, a thermo-catalytic method that synthesizes ammonia from hydrogen and nitrogen gas streams in a stoichiometric ratio. Hydrogen and nitrogen become available at around 30 bar, and this is the pressure level at which the synthesis gas (syngas) is produced. Subsequently, a compression station must pressurize the syngas (PR 5–7) to shift the chemical equilibrium towards the right side, resulting in ammonia. This task becomes more challenging with centrifugal compressors required for large production volumes.
The typical – and well proven – configuration in syngas service involves two units on the sides of a driver. When transitioning to green ammonia, the mixing pressure to produce the syngas is flexible, introducing a degree of freedom in the synthesis loop configuration. However, reducing the effort in compressing hydrogen means demanding more PR for the syngas portion, and vice versa. This flexibility leads to a ‘short blanket’ dilemma, where the limits of standard centrifugal compression technology hinder any increase in synthesis gas PR without adding compressor units.
To fully benefit from this additional degree of freedom, a technology capable of unlocking such PR is needed, as well as being capable of successfully addressing hydrogen compression duty challenges. Applying HPRC technology to syngas (75% hydrogen in volume) dramatically increases PR capability. This allows for maintaining a two-compressor configuration, achieving an unparalleled PR beyond 25: a perfect fit for the desired green ammonia process optimization of front-end/back-end pressures.
The adoption of HPRC for green ammonia production results in tangible benefits in both CAPEX and OPEX, especially in large production volumes.
Compressors for hydrogen storage
To reduce the footprint despite hydrogen’s low volumetric energy density, the typical storage pressure for hydrogen is not less than 200 bar and generally below 350–400 bar. When considering green hydrogen production via electrolysis, the hydrogen storage compression system should be capable of compressing hydrogen from an ambient/30 bar suction pressure up to the storage pressure range of 200–350 bar (PR from 6.6 to 350). Depending on the scale of green hydrogen production, several tons per day of hydrogen feed the hydrogen storage, varying based on energy source fluctuations and production profiles.
Due to high capacity and high PR requirements, reciprocating and diaphragm compressors are the most viable solutions. To maintain gas purity, no piston lubricating cylinders are allowed, and capacity control from 20% to 100% is often required. Hydrogen-rich dry piston compressors are well-referenced in hydrogen and ammonia refinery and petrochemical applications. Typical applications range from a few TPD to 150 TPD in a single machine, with variable suction and delivery pressure up to 100 bar.
To reach storage pressures, high-pressure dry tandem cylinders have been developed specifically for the hydrogen market. In Figure 4, two typical tandem cylinder arrangements are shown for low- to medium-pressure cylinders and for high-pressure levels. The tandem configuration allows for more compression stages in a single cylinder, and the arrangement of the high-pressure stage with respect to the packing ring reduces the pressure drop on the packing rings, enabling the system to reach 350 bar with practically the same referenced sealing technology.
High-pressure tandem cylinders can be designed to be arranged on existing API 618 reciprocating compressor frames, which are well-referenced for hydrogen and high-load service. Typically, on each standard frame, up to 10 cylinders (standard double effect or tandem) can be installed to satisfy the required high PR and high capacity. Considering a suction pressure of 19 bar and a discharge pressure of 350 bar, a Baker Hughes frame of medium capacity can handle up to 48 TPD in a single machine.
Diaphragm compressors are limited by diaphragm size to the order of the TPD. To serve high capacity, several diaphragms are needed to cover a single piston compressor duty. Diaphragm compressor costs are growing linearly depending on capacity while piston compressor costs are mainly driven by frame size and cylinder number, with lower impact on overall cost. As shown in Figure 5, piston compressor CAPEX does not grow directly with capacity as diaphragm compressor. At 12 TPD, a single high flow piston compressor with a suction of 19 bar and a delivery pressure of 350 bar can save up to 70% CAPEX compared to an equivalent diaphragm solution (12 machines).
Reliability and availability of API 618 piston compressors are very high compared with diaphragms, which require redundancy.
Compressors for hydrogen refueling stations
The share of hydrogen in mobility is negligible today, and current hydrogen mobility infrastructures need to scale up by several factors to meet the net-zero scenario. Low-carbon hydrogen is mainly produced by electrolysis and requires compression for transportation, storage, and utilization. Considering the trade-off between tank costs and range, the majority of heavy-mobility trucks are equipped with 350 bar systems, while the standard for light mobility is 700 bar.
The share of hydrogen in light-duty mobility is strongly impacted by the adoption and development of fully electric alternatives. In Figure 6, the average capacity of low-carbon hydrogen projects related to mobility at FID or feasibility level is plotted per the expected year of production start, together with the total number of projects. The average production capacity is increasing from 7 tons/year in 2021 to more than 100 tons/year in 2027.
In this evolving market, Baker Hughes and other OEMs in the API 618 reciprocating compressor market are evolving their solutions toward heavy-duty mobility to serve the 350-bar market with compact solutions with vertical dry cylinders capable of compressiing hydrogen from 30 to 550 bar, with a capacity ranging between 2 and 10 TPD, depending on the number of installed cylinders. Typically, mobility solutions are designed with a vertical cylinder configuration and are on-skid to minimize installation activities. Vertical cylinders allow for a compact footprint, and this configuration is preferred over the standard API 618 horizontal cylinder configuration.
Final remarks
In this article, an overview of available compression and expansion technologies has been presented, drawing on the Baker Hughes portfolio, recent experiences, case studies in the field of energy transition, and planned developments. The aim was not only to introduce the technical aspects but also to offer insights into the economic fits and gaps of current technologies to cope with the forecasted increase in demand for hydrogen production.
As a general conclusion, there is no single ‘best-fit’ technology for hydrogen compression; rather, multiple complementary technologies allow the configuration of the best solution for each specific case, taking into account the operating parameters. Therefore, careful attention should be given to the compression configuration, in conjunction with process/plant design, to enable the highest optimization from both technical and economic standpoints.
