Molten-Salt Pumps: Why Temperature Decides the Machine, Not Pressure

A molten-salt pump moves a fluid that solidifies if it is allowed to cool, and the standards record shows which half of the problem was settled first. China's national standard for the working fluid, GB/T 44800-2024, was published on 26 October 2024 with normative annexes for melting point, decomposition temperature, specific heat, density and corrosion testing. The national standard for the pump itself — a technical specification for molten-salt pumps in solar thermal power stations — only entered the plan on 30 July 2026, as project 20264103-T-604, with a 15-month drafting period. The fluid came first; the machine is being written now.

This article sets out why the fluid and not the pressure sets this duty, what the machine has to do that an ordinary hot-fluid pump does not, who has put what on the record, and where that record stops.

Why did the standard for the fluid arrive before the standard for the pump?

Because the fluid is a substance that can be specified by its own properties, while the pump is a machine whose requirement only becomes concrete once the system around it is defined — and China's standards system has published the first and is only now drafting the second.

The fluid standard is GB/T 44800-2024, published and effective on 26 October 2024. Its normative annexes are the clearest available description of what a molten-salt duty is: melting point, decomposition temperature, specific heat, density and a corrosion performance test are not commentary, they are appendices a supplier is measured against. Two of those five properties — melting point and decomposition temperature — are a lower and an upper bound on the fluid, and the machine lives in the gap between them.

The pump appears next, at first as a clause inside a system standard rather than as a product in its own right. GB/T 45313-2025, published on 28 February 2025 and effective from 1 September 2025, is the thermal-storage system standard, and its table of contents gives the pump a section of its own, §5.4 on molten-salt pumps. An association standard of 25 December 2024, T/CIEP 0208-2024, carries the same arrangement and goes further, with a numbered clause for each of the three pump duties in the system.

Then the pump gets a standard of its own. National standard project 20264103-T-604, a technical specification for molten-salt pumps in solar thermal power stations, was placed on the plan on 30 July 2026 with a 15-month drafting period, under the pump standardisation committee TC211 and the China Machinery Industry Federation, drafted by Zhejiang Kesheng Technology (浙江可胜技术股份有限公司), Jiangsu Feiyue Pump Industry (江苏飞跃泵业股份有限公司) and the Shenyang Pump Research Institute (沈阳水泵研究所有限公司). Its scope is specific rather than general: high-temperature molten-salt vertical suspended pumps for solar thermal power stations, with other high-temperature molten salts — chloride salts and molten alkalis — allowed to reference it.

Read down that table and a sequence appears: the medium, then the system, then — only from mid-2026 — the machine. The fourth row says the most about the duty. Freeze prevention and insulation have a design guideline of their own, which puts the risk of the salt setting solid in the machine on the same footing as the risk in the pipework.

What does the machine have to do that an ordinary hot-fluid pump does not?

It has to hold the fluid inside a temperature window, empty itself completely when it stops, and carry a long vertical shaft — and its bearings — through the hot salt.

The association standard T/CIEP 0208-2024 sets the duty out as three machines rather than one. A cold-salt pump feeds low-temperature molten salt to the collection system; a hot-salt pump feeds high-temperature molten salt to the heat-exchanger system; an attemperating pump adjusts the salt temperature at the inlet to that heat exchanger. They are three functions at three temperatures in one circuit, and a station can need all three.

The rest of the same clause reads as a list of what goes wrong in service.

The 30,000-cycle figure is the one that turns a reliability wish into a design requirement. It is the order of magnitude of daily start-stop cycles over a station's working life, and it identifies the failure mode the specification is aimed at. Not abrasion, and not corrosion: the repeated thermal excursion of a long shaft, a bearing and a seal.

What actually has to be solved, once the fluid is hot?

Four items, and none of them is pressure: the hydraulic model at temperature, the corrosion-resistant material, the bearing that runs hot, and the seal at the top of a long shaft.

Two public lists name the same development programme, and reading them side by side is more informative than either list alone.

The industry association's account, 2026. The pump branch of the China General Machinery Industry Association (中国通用机械工业协会泵业分会) reports four key technologies solved: the hydraulic model for high-temperature molten salt, corrosion-resistant special materials, long-term lubrication of the high-temperature bearing, and leak-free sealing.

The municipal state-asset commission's account, 2022. Four years earlier, the Lanzhou municipal State-owned Assets Supervision and Administration Commission (兰州市人民政府国有资产监督管理委员会) recorded the same programme's solved technical problems as hydraulic performance, material corrosion, bearing lubrication, shaft-system balance, bearing sealing and a high-temperature thermal shield.

The overlap is the finding. Both lists name the hydraulic model, the material, the bearing and the seal — and the 2022 list adds two that belong to a long vertical machine rather than to a hot fluid as such: shaft-system balance, and a thermal shield between the hot end and everything above it. That is a machine-specific list, and it is the reason the standard pays as much attention to vibration and to bearing temperature as to flow.

The two records also bracket the same development: in July 2022 the machine had completed a cold-state test on water and was about to start a four-month high-temperature test; in August 2026 the same programme is reported in engineering application. The 2022 record adds that the work was supported by the Chinese Academy of Sciences' Strategic Priority Research Programme on Transformative Clean Energy (变革性洁净能源先导专项).

Put plainly: on this duty the wetted metal is a materials problem, and everything else is a thermal problem wearing a mechanical costume.

Who has put what on the record?

Three makers, with published duties at 595 °C, 650 °C and 700 °C — and the records do not carry the same weight.

The 595 °C machine, with the longest service statement. Jinan Huawei Pump Industry (济南华威泵业有限公司) supplied one high-temperature, long-shaft, large-flow hot-salt pump for the 100 MW molten-salt tower CSP station of Shouhang Hi-Tech (首航高科) at Dunhuang, described in the supplier's account as China's first 100 MW-class national solar thermal power demonstration station, with an 11-hour molten-salt storage system and 24-hour continuous generation. The design values are published as 595 °C, 662 m³/h, a head of 41.8 m and a submerged depth of 18 m. It was commissioned on 6 December 2021 and accepted on site as the main pump; as of August 2022 it had run reliably for nine months, including 262 continuous hours in June 2022. The same maker had earlier supplied one long-shaft attemperating pump for the 50 MW Hami tower CSP project, commissioned in December 2019.

The 650 °C target. Jiangsu Feiyue Pump Industry has the longest published development record. Its molten-salt work began in 2016, in a project on key technologies for high-reliability molten-salt pumps under extreme high-temperature duty run with the Institute of Metal Research of the Chinese Academy of Sciences and Jiangsu University. In December 2019 a 17 m hot-salt pump of its manufacture started and passed commissioning in a single attempt at the 50 MW molten-salt linear-Fresnel CSP project of Lanzhou Dacheng (兰州大成) at Dunhuang. In January 2021 a high-head cold-salt pump it supplied to the 50 MW Delingha tower CSP station was installed and commissioned. In 2025 a follow-on project on key technologies for a next-generation ultra-high-temperature molten-salt thermal storage system entered the Jiangsu provincial frontier technology research and development programme with provincial funding, aimed at a 650 °C, 12-hour storage demonstration. The company is also one of the three drafting units of the national pump standard.

The 700 °C class. Lanzhou Lanpump (兰州兰泵有限公司), part of Jincheng Industrial Group (金城实业集团), developed the machine with the Shanghai Institute of Applied Physics of the Chinese Academy of Sciences and Lanzhou University of Technology. The industry association reports the result as the world's first engineering-scale 700 °C ultra-high-temperature submerged long-shaft molten-salt pump, now applied at the molten-salt storage centre project of the Huaneng Xi'an Thermal Power Research Institute (华能西安热工研究院) — rated at 500 MW / 2,000 MWh and described in the same report as the world's largest electric heat-storage molten-salt project. The programme also built what the association describes as the world's first 700 °C molten-salt pump test rig.

Three things stand out from that table. The highest-temperature machine does not have the longest service record — the nine-month statement belongs to the 595 °C machine, while the 700 °C machine is reported as an engineering application. The 650 °C figure is a funded target, not a delivered temperature. And the makers divide by duty rather than by size: a hot-salt pump, a cold-salt pump and an attemperating pump are three different jobs in the same station, which is why the standard lists them separately.

Where does this route stop?

At five places, all of them visible in the published record.

The temperature figures are not one curve. 595 °C has nine months of service behind it. 650 °C is a funded project target. 700 °C is reported as an engineering application. Reading the three as a single capability curve would merge a machine that has run, a machine that is funded to be built, and a machine reported to have entered service.

The longest service statement is not on the newest machine. Nine months of operation and 262 continuous hours is the only multi-month service record in this set, and it belongs to the 595 °C machine on a 100 MW CSP station. The 700 °C machine's record is an application announcement. The two should not be quoted as one result, and neither establishes a maintenance interval.

No life figures are published. There is no mean time between failures, no overhaul interval, no measured start-stop cycle count against the standard's 30,000-cycle requirement, and no bearing or seal life in hours for any machine in this article. The 30,000 figure is a requirement written into a standard; it is not a measured result.

Two Dunhuang projects are two different machines. The 50 MW molten-salt linear-Fresnel station that took a 17 m hot-salt pump in December 2019, and the 100 MW molten-salt tower station that took a long-shaft hot-salt pump in December 2021, sit in the same place and are often referred to as one Dunhuang project. They are different owners, different collector technologies, different storage durations and different pumps.

The pump standard is a plan, not a published standard. 20264103-T-604 was placed on the plan on 30 July 2026 with a 15-month drafting period. Until it is published there is no national product standard for this machine class, and every requirement quoted in this article comes from a system-level standard or an association standard. And the 700 °C machine's record — the test rig, the four-year programme, the named technologies and the engineering application — is reported by its developer, its industry association, and the municipal state-asset commission that owns the group. There is no independent performance benchmark of that machine in the published record.

What does the trend add up to?

Three movements are visible, and they sit at different levels of maturity.

The temperature ceiling moved from a delivered 595 °C duty to a 700 °C engineering-application class, and the route between them is recorded in stages rather than as one leap: a cold-state test on water, then a four-month high-temperature test, then engineering application — the first two recorded in 2022, the third in 2026.

The standardisation moved up the chain from the fluid to the machine. The medium was specified in October 2024, the storage system — with the pump as a clause inside it — in 2025, and the pump as a product class entered the national plan in July 2026. That order is the significant part: for this machine class, the substance being pumped has been defined for longer than the machine that pumps it.

And the design problem list moved away from the wetted metal. The named technologies are the hydraulic model, the material, the bearing and the seal, with shaft-system balance and a thermal shield alongside them — a list about heat and vibration rather than about corrosion alone.

What has not moved is the evidence type. Every figure in this article is a standard requirement, a design value or a developer's statement, and the longest measured service window is nine months.

What this article does not cover

Note that this article does not select a pump, size a machine, calculate head or net positive suction head, specify a material, a bearing, a seal or an insulation system, or give prices and delivery times. It describes the duty, the standards that define it, the published project records and the boundary of that evidence. It is not a purchase specification and it carries no compliance determination.

Where the next constraint sits

The constraint that follows this one is the same fluid seen from the other side. Molten salt has to be kept hot to stay liquid at all, and Liquid Hydrogen Pumps: What the −253 °C Duty Demands works the opposite case — a fluid that has to be kept cold, where one heat leak turns it into gas and the machine loses its suction.

The other adjacent duty changes the reason the fluid misbehaves. Instead of a fluid with a melting point, CO₂ Injection Pumps: Why the Clearance Volume Decides the Efficiency takes a fluid whose density moves steeply with pressure and temperature, and shows why the volume left in the valve chamber sets the efficiency.

Sources