Liquid Helium Ballast Refrigeration

Figure 1a: Gravimeter dewar. Credit: R. Fagaly 1b: Alternate design to minimize liquid helium. Credit: R. FagalyThe Need for Vibration-Free Cryogenics The measurement of local gravitational fields requires continuous, high precision and extremely stable measurements. In 1968, the development of the superconducting gravity meter (SG)[1] vastly improved the state of the art. In contrast to mechanical quartz or metal springs, SGs use a superconducting sphere levitated in an ultra-stable magnetic field generated by persistent currents in a pair of superconducting coils. This cryogenic design enabled sensitivities that can exceed 10-10 m/sec2 (i.e., 0.01 ppb of the earth’s gravitational field) and drifts less than 60 nm/sec2/year. Many gravity measurements (hydrology, volcanology, geothermal energy and postglacial rebound) require durations that last years, and, in many cases, decades. To fulfill this need, extremely long hold time cryostats were developed to maintain the SG at operational temperatures close to 4 K. Special techniques were also developed to prevent cryocooler-induced accelerations from contaminating the geophysical signals of interest. While most superconducting quantum interference device (SQUID) measurement systems[2] do not require multiyear measurement times, they benefit from improved cryogenic systems that minimize or eliminate the expenditure of liquid helium.

Cryocooled Systems
Early SGs and SQUID systems were operated in carefully shielded and evacuated dewars with reservoirs of liquid helium that had to be replenished periodically by transferring liquid from storage dewars. Hold times between transfers varied dramatically depending on the dewar storage volume and heat loads into the helium reservoir. Early SG dewars had volumes of 200 liters and five-inch diameter necks that allowed the insertion of the sensor into the dewar. Typically, 100 liters of liquid would be transferred every 20 to 30 days to maintain the 4 K liquid helium reservoir. This process not only disturbed the gravity record but was expensive. The earliest SQUID dewars ranged from five to 25 liters and were of fiberglass construction[2] to allow measurements of external magnetic fields; these required refills ranging from twice/week to twice/month under normal operating conditions.

As an alternative to the use of liquid cryogens, closed-cycle refrigeration[3] became desirable for several reasons. These include reduction of operating costs, simplified use in remote locations, avoiding interruptions in cryogen deliveries, safety and the convenience of not having to transfer at periodic intervals. Unfortunately, directly connecting the experiment to the cryocooler introduced significant noise.

The first practical cryocooled superconducting sensor operating at 4 K (as opposed to the use of cryocoolers for MRI systems where vibration- and motion-induced noise was not significant) was the BTi CryoSQUID.[4] Based on a two-stage Gifford-McMahon (GM) refrigerator with 1 W refrigeration power at 15 K, the use of a Joule-Thompson (JT) stage allowed 4 K operation with reduced vibration. The subsequent development of pulse tube refrigeration[5] allowed operation with significantly reduced vibration. However, there remained three significant obstacles to using closed-cycle refrigeration with highly sensitive superconducting instrumentation:
• Mechanical movement of the internal components of the cryocooler,
• Short-term temperature variation of the cryocooler’s coldhead, which can vary from ±0.1 K to ±0.5 K at Hz scale frequencies, and
• Long-term temperature changes of the cryocooler’s coldhead as the unit ages.
Liquid Helium Ballast Systems

To minimize these obstacles, a mechanically isolated cryocooler was developed first to extend the hold time of the liquid helium reservoir, and eventually to liquefy helium gas to both fill and maintain the liquid helium reservoir. Figure 1a and 1b show two different designs that have been implemented. Figure 1a shows the cryocooler inserted in the neck of the dewar, either partway or into the liquid ballast tank. Figure 1b shows the cryocooler located in the vacuum space with flexible connections (typically copper braid), thermally coupling the two cold stages of the coldhead to the radiation shields of the dewar. This design, suitable for a variety of SQUID magnetometer systems, minimizes the amount of liquid helium (typically 1 ~ 2 liters), while allowing interruptions of many hours for either movement of the magnetometer system or ultraprecise measurements.

In 1981, long before the development of 4 K GM cryocoolers,[3] GWR Instruments, in conjunction with S.H.E. Corporation, developed dewar systems (Model TT40, TT60 and TT70) based on the Figure 1b design that used a doubly shielded 200-liter dewar in conjunction with a two-stage (APD Cryogenics DE202) GM cryocooler capable of achieving temperatures below 10 K. The upper stage was used to cool the outer thermal shield, and the lower (cold) stage was used to cool the inner stage. The use of the cryocooler reduced the liquid helium boiloff of the dewar from 1 liter/day to 0.2 liter/day, allowing a 1,000-day hold time. Connecting the cryocooler to the shields via flexible copper braiding significantly reduced the coupling of the vibration generated by the cryocooler to the sensor. Although this dewar had an exceptional hold time, only a few more were manufactured since it was very difficult to remove the coldhead for servicing at the required one-to-two-year intervals. The next models were similar to that shown in Figure 1a, but with the sensor first inserted through the neck of the dewar, and the coldhead inserted afterward. Helium gas in contact with concentric heat exchangers transferred heat from the dewar shields to the coldhead. A separate support frame supported the coldhead without allowing contact with the neck. This design entirely isolated the coldhead vibrations from the dewar and gravity sensor and was a major advance in decreasing this source of noise on the sensor. This design simplified coldhead service and decreased noise, but the tradeoff was a decrease in hold times to less than 500 days.

In 1993, GWR produced a 125-liter compact dewar design that was much smaller and one that could be operated at many preexisting geodetic observatories. It used the same APD Cryogenics DE202 coldhead; however, the gravity meter was installed permanently into the dewar belly through removable seals in the dewar belly and vacuum chamber. This allowed the neck diameter to be reduced from 12.7 cm to the diameter of the coldhead and made it both easier to transfer heat from the coldhead stages to the dewar shields and to isolate the coldhead vibrations from the dewar and sensor. After testing at Royal Observatory in Brussels, on August 4, 1995, the first compact SG C021 was moved to a room at the end of a 100-meter-long tunnel at a seismic station in Membach, Belgium, where it is still operating today. This instrument is testament to the design of a liquid helium ballast system. At the end of 2022, this gravimeter had measured variations in gravity for 10,000 consecutive days (over 27 years), and it holds two world records: first, for measuring a continuous gravity signal at a single location for the longest time in the world, and second, for the longest circulation of electric currents in superconducting circuits.[6]

By 1997, Leybold Vacuum introduced the first practical closed cycle 4 K refrigeration system, the Leybold KelKool 4.2. Its GM coldhead cooled below the 4.2 liquefaction temperature, producing about 50 W cooling power at its upper stage at 50 K and about 0.5 W cooling power at the lower stage at 4 K. Soon afterward, GWR produced an ultralong dewar based upon the 125-liter compact design but with its neck enlarged to accommodate the much larger 4.2 GM coldhead. It is believed that this is the first design that coupled a 4 K GM coldhead with a dewar and a liquid helium storage volume in which the coldhead was used to liquefy He gas into a storage volume cryostat using only the GM cryocooler.[7] Although very reliable, the coldhead was too heavy for one person to handle and required a support crane for insertion and removal. In addition, the compressor and water chiller required 7 kW of power versus the 2-kW power for the APA DE202 202 system. Two systems were manufactured with the Leybold KelKool 4.2 and two with the Leybold 4.2 Lab coldhead, which reduced the power load to 3.5 kW. Soon after the last shipment to the Japanese Showa Station in Antarctica in March 2003, Leybold discontinued manufacturing these cryocoolers.

At about the same time, Sumitomo Heavy Industries (SHI) entered the market with a new smaller 4 K refrigeration system, the SHI RDK-101 coldhead and CAN-11 compressor. This system produces about 0.2 W of cooling power at 4 K and is about the same size as the APD 202 DE202 so that it can easily be handled by one person. This led to the development of GWR’s present iGrav superconducting gravimeter (Figure 2) that uses a 16-liter dewar with the coldhead supported by a small external frame with helium gas exchanging heat between the coldhead and the dewar shield. Helium gas is sealed in the dewar using a flexible membrane, and heat is added to the dewar’s neck to control the pressure close to atmospheric pressure. Since the pressure differential is very small across the sealing membrane, the temperature variations of the storage volume vary with atmospheric pressure with a dependence of about 1 mK/mBar. The development of the iGrav’s dewar/refrigeration system is the result of about 25 years of development, and to date, about 50 of the iGrav systems have been shipped worldwide.

The liquid helium ballast systems minimize the obstacles to using closed-cycle refrigeration for operating highly sensitive superconducting instrumentation. The liquid ballast technique isolates the experimental volume from the coldhead vibrations, and, because of the large heat capacity of the liquid helium, reduces the Hz level ±0.1K (or greater) thermal variations of the cryocooler to mK levels. It also allows for short interruptions of electrical power without system warm-up, essentially acting as a thermal storage battery. “Battery life” depends critically on the dewar volume and design of the interface of the coldhead and dewar neck. GWR’s iGrav, with its reduced size and weight, simplifies installation and transportation while providing more than 10 days of “battery life.” This allows ample time for most service requirements and allows the instrument to be moved from one location to another without warm-up.

Another major advantage of liquid helium ballast systems is that liquid helium is not needed for the initial cooldown. By injecting room temperature gas into the cryogenic insert, the cryocooler is used to condense gas directly into the reservoir. Yet another advantage of a liquid helium ballast system is that, if cryocooler vibration interferes with the experiment, it can be shut down during the measurement time. This is particularly attractive for SQUID measurements or where portability is needed.

Initially developed for GWR’s superconducting gravity meters[7],[8] in conjunction with Tristan Technologies, liquid helium ballasted systems have been implemented on commercial SQUID neuromagnetometers,[8],[9] rock magnetometers and magnetic microscopes. This method to liquefy and maintain helium led directly to small-scale liquefaction systems[9] to produce and maintain liquid helium on demand for general use. These systems are ideal for helium gas recovery and reuse, which is important for this non-renewable resource. This product is currently licensed to QD-USA[10] for manufacture and distribution worldwide.

GWR recently delivered its one hundredth cryocooled system. These superconducting gravity meter systems are often operated in rugged field conditions for decades without interruption. In many cases, this “helium battery” concept provides a superior solution for bridging power interruptions compared to power backup systems using generators and batteries. For information on GWR’s superconducting gravity systems or liquid helium ballast systems, contact [email protected]. For liquefaction systems, contact [email protected].

The late Ray Sarwinski was instrumental in the early development of cryocooled dewars, both at SHE and Tristan Technologies.
*GWR Instruments, San Diego, CA 92121
*Tristan Technologies, San Diego, CA 92121

[1] Hinderer J., Crossley D. and Warburton R.J, “Superconducting Gravimetry” in: Gerald Schubert, ed., Treatise on Geophysics, 2nd edition, vol 3. (Oxford: Elsevier, Amsterdam, 2015) pp. 59-115.
[2] Fagaly, R.L., “Superconducting quantum interference device instruments and applications”, The Review of Scientific Instruments, 77, 101101 (2006)
[3] Walker, G, Miniature Refrigerators for Cryogenic Sensors and Cold Electronics, Oxford: Clarendon Press, 1989
[4] Buchanan, D.S., Paulson, D.N., and Williamson, S.J., “Instrumentation for clinical applications of neuromagnetism,” in Advances in Cryogenic Engineering, 33, pp. 97-106, R. W. Fast, (ed.), New York: Plenum, 1988
[5] Heiden, C., “Pulse tube refrigerators: a cooling option,” in SQUID Sensors: Fundamentals, Fabrications and Applications, pp. 289-306, Weinstock, H. (ed.), Dordecht/Boston/London: Kluwer Academic Publishers, 1997
[8] Okada, Y., Pratt, K., Atwood, C., Mascarenas, A., Reineman, R., Nurminen, J., and Paulson, D., “BabySQUID: a mobile, high-resolution multichannel MEG system for neonatal brain assessment”, Review of Scientific Instruments, 77, 024031 (2006)
[9] US patents 20130047632-A1 (2013) and 20130192273-A1 (2013)

Image: Figure 1a: Gravimeter dewar. Credit: R. Fagaly 1b: Alternate design to minimize liquid helium. Credit: R. Fagaly

Share this post:

Comments on "Liquid Helium Ballast Refrigeration"

Comments 0-5 of 0

Please login to comment