DOE has prepared a Multi-Year Research, Development, and Demonstration Plan to provide hydrogen as a viable fuel for transportation after 2020, in order to reduce the consumption of limited fossil fuels in the transportation industry. Hydrogen fuel can be derived from a variety of renewable energy sources and has a very high BTU energy content per kg, equivalent to the BTU content in a gallon of gasoline.
The least expensive delivery option for hydrogen involves pipeline transport of the hydrogen from the production sites to the population centers where the vehicles will see the highest demand and thus have the greatest impact on reducing the US dependency on fossil fuel. The cost to deliver the hydrogen resource to the refueling stations will add to the ultimate cost per kg or per gallon equivalent that needs to be charged for the hydrogen fuel. In order to keep the cost per gallon (equivalent) of hydrogen fuel as low as possible to the consumer, it is necessary to reduce the cost of its transmission to the refueling station. This can be done by reducing the cost of the compressors, as well as all other refueling station components, while also increasing their energy efficiency. DOE has set a target of $1/GGE (Gallon of Gasoline Equivalent).
To meet these goals, the DOE has commissioned Concepts NREC (CN) with the project entitled: “The Development of a Centrifugal Hydrogen Pipeline Gas Compressor”. CN’s engineering project plan consists of three distinct phases, as detailed in Figure 1. Phase I of this project has been completed and is described in this article.
The engineering approach used by CN to accomplish these goals is to utilize state-of-the-art aerodynamic/structural analyses to develop a high-performance centrifugal compressor system for pipeline service. The centrifugal-type compressor is able to provide high pressure ratios under acceptable material stresses for relatively high capacities (flow rates that are not usually higher than what a piston compressor can provide). CN’s technical approach also includes the decision to utilize commercially available, and thus proven, bearings and seal technology to reduce developmental risk and increase system reliability at a competitive cost. CN has researched the use of a material that is compatible with hydrogen and that can enable the highest possible impeller tip speeds, which results in the reduction of the number of required stages while meeting DOE’s goals for overall pressure ratio and efficiency.
In order to minimize the development time and ensure industrial acceptance of the design for the new pipeline compressor system, CN has assembled a project team to assist in the advanced engineering of the compressor while also preparing an implementation plan that can provide for near-term industrial pipeline applications. The team consists of Praxair (to provide industrial user experience and “hands-on” experience), Texas A&M University (to provide materials science expertise and testing), and HyGen Industries (to provide experience in hydrogen fueling infrastructure).
The engineering challenge to implement this technical approach is to design a compressor stage that can achieve the highest acceptable pressure ratio and thermodynamic efficiency per stage, but also by using as few stages as possible and the smallest diameter. For centrifugal-type compressors, the pressure ratio is proportional to rpm2 and radius2. Thus, even a small increase in tip speed or impeller radius results in significant increases in pressure. The aerodynamic design challenge in reducing the number of stages is to maximize the tip speed of the centrifugal compressor impeller, while also staying within acceptable design safety levels of the strength limitations of material and utilizing advanced diffuser systems to maximize recovery of dynamic head into static pressure. However, material stresses also increase proportional to the square of the rotor speed (rpm2) and radius (radius2) and also by material density. Ultimately, the major constraint is imposed by the limitations of the maximum stress capability of the impeller material. This constraint is further aggravated by the need for the material selection to consider the effects of hydrogen embrittlement on the strength of the material.
These engineering challenges are best summarized by their interdependencies, as illustrated in Figure 2. As may be observed from Figure 2, high impeller tip speeds will enable high pressure ratios to be attained with fewer stages. Thus, a major challenge for the project has been to identify the material that can enable the highest tip speeds to be attained, while also sustaining the stresses that are imposed by these tip speeds. CN has met all of these engineering challenges in order to provide a pipeline compressor system that meets DOE’s specifications.
A summary of the results and accomplishments of this project to date are given in Figure 3 for each of the engineering specifications and objectives as set by DOE for the project.
In order to achieve these accomplishments, CN developed several computer design models that would optimize the design choice. These models include:
- Compressor Package Performance Model that provides a single point summary of each of the components within the package.
- Cost Model using algorithms to determine the relative component cost and operational risks associated with compressor design specifications.
- Engineering Reliability and Maintenance Cost Model that uses a consistent methodology and algorithms to determine the relative reliability and maintenance cost for a piston and centrifugal compressor pipeline package.
The Cost and Performance Model enabled the analysis of over 30 combinations of centrifugal compressor impeller speeds, the number of stages, with a single or dual impeller-shaft design using a one or two-step gearbox, with a high- or low-speed prime mover drive arrangement. The best choice, with respect to conformity to commercially available system specifications along with high efficiency and low operational risk, is highlighted. This choice is a single, overhung (cantilevered) impeller attached to a drive shaft that includes a shaft seal, bearing, and drive pinion. The impeller rotor is designed without a bored-hub in order to reduce the hub “hoop” stresses. This requires the impeller to be mechanically attached to the high strength steel alloy drive shaft with a patented design attachment system that enables the rotor to be removed from the gearbox without removing the drive shaft and thus without disturbing the shaft seal and bearings.
The compressor selection uses six stages, each operating at 60,000 rpm with a tip speed of less than 2,100 ft/s. Each compressor, shown in Figure 4, is 8 inches in dia. and has an overall stage efficiency of between 79.5 and 81%, for an overall compressor efficiency of 80.6%.
Materials Selection Validation
The material chosen for the compressor is an aluminum alloy. The choice is based on its mechanical strength-to-density ratio or (Syield/ρ), which can be shown to be a characteristic of the material’s ability to withstand centrifugal forces. Several grades of aluminum have a strength-to-density ratio that is similar to titanium and high strength steels at the 140°F (max) operating temperatures that will be experienced by the hydrogen compressor by providing intercoolers between each of the six stages. However, unlike titanium and most steels, aluminum is recognized by the industry as being very compatible with hydrogen, not showing any susceptibility for hydrogen embrittlement. Aluminum also helps to reduce the weight of the rotor, which leads to an improved rotor dynamic stability at the 60,000 operating speed. A rotor stability and critical speed analysis has confirmed that the overhung design is viable.
The project team includes researchers at Texas A&M (Dr. Hong Liang and Graduate Assistant Matt Sanders) who are collaborating with CN to confirm the viability of aluminum alloys for this compressor application. A Test Protocol has been established based on a series of discussions with notable researchers in several national laboratories including:
- Sandia National Labs (Fracture Mechanics Testing; Dr. Chris San Marchi)
- Savannah River National Labs (Specimen “Charging” with Hydrogen Plus Tensile Testing with H2; Dr. Andrew Duncan)
- Argonne National Labs (Dr. George Fenske)
In order to quantify the susceptibility of the aluminum and titanium alloys to hydrogen embrittlement, Texas A&M has fabricated a special test fixture identified as a Small Hole Punch assembly to be used with a standard Instron tensile test machine. The test apparatus will puncture a hole into a test specimen that measures 3 mm in diameter and 0.5 mm thick. The purpose of these tests is to determine the relative effects of hydrogen exposure on the metal matrix, and thus, determine if the metal strength has been compromised.
The complete hydrogen pipeline compressor package that integrates all of these major components is shown in Figure 5. The complete modular package, as shown, is 26 ft long x 10 ft tall x 6 ft wide (at the base) x 8 ft. wide at the control panel, approximately ½ the footprint of a piston-type, hydrogen compressor. The packaged module can be transported to the installation site as a pre-assembled package with a minimum of final alignment, water piping, and electrical power connections, instrumentation, and controls start-up.
Preliminary design details of the completed pipeline compressor module include:
- Compressor design conditions confirmed by project collaborators
- Pinlet= 350 psig, Poutlet=1,285 psig; Flow rate= 240,000 kg/day
- 6-stage, 60,000 rpm, 3.56 pressure ratio compressor
- Integral gearbox pinions driving individual, overhung impellers
- Design of compressor’s major mechanical elements completed and satisfied by two manufacturers per component
- KMC tilting pad radial and thrust bearings designs validated for use
- FlowServe single, gas face-seals have been validated for use
- Heat exchanger specifications met by two manufacturers to cool hydrogen gas to 100°F between stages
- Tranter Plate-type Heat Exchanger Design
- Heatric Heat Exchanger (compact, plate-fin surface core)
Conclusions and Future Directions
The preliminary engineering and design of an advanced pipeline compressor system has been completed that meets DOE’s performance goals for a reliable, hydrogen pipeline compressor system, with a footprint one-half the size of existing industrial systems and at a projected system cost of approximately 75% of DOE’s target and a maintenance cost that is less than the $0.005/kwh. The advanced centrifugal compressor-based system can provide 240,000 kg/day of hydrogen from 350 to 1,300 psig high for pipeline-grade service. This has been accomplished by utilizing state-of-the-art aerodynamic and structural analyses of the centrifugal compressor impeller to provide high pressure ratios under acceptable material stresses. The technical approach that has been successfully implemented to reduce the developmental risk and increase the system reliability while maintaining a competitive cost includes using commercially available, and thus proven, bearings and shaft seal technology, as well as acceptable practice in bearing and gear loading.
The resultant design provides a compressor that meets DOE’s Hydrogen Plan for future pipeline delivery, as well as for use by the industrial, hydrogen gas industry where there are presently 1,200 miles of pipelines providing 9 million tons per year of hydrogen gas for industrial process chemical applications. A collaborative team has been assembled consisting of Praxair, Texas A&M (a materials researcher), and HyGen Industries (a hydrogen refueling industry consultant).
Future efforts include:
•Phase II. Detailed Design (01/2010 to 08/2010)
- Detailed subsystems modeling
- Detailed integrated systems analysis
- Critical components design, testing and development
•Phase III. System validation testing (09/2010 to 06/2011)
- Component procurement
- Two-stage centrifugal compressor system assembly and lab test
CN engineering personnel who have contributed to this article include Jamin Bitter, Kerry Oliphant, Sharon Wight, Glenn Derbyshire, Ken Heidelmeier, Ron Provencher, Fred Becker, Dr. Karl Wygant, and Dr. Louis Larosiliere.