D1F.1 Hydrogen compatibility of ball valves from the natural gas transmission grid

This report describes the execution and results of work package WP1F (Testing valves in the transmission grid), which is a part the HyDelta research program. The research exists of measurements on the external leakages of ten ball valves with both natural gas and hydrogen at pressures above 16 bar. Additionally the internal leakage of twelve ball valves and external leakage of seven plug valves in a natural gas grid were measured.

The goal of the measurements with both natural gas and hydrogen is to find a relation or ratio of the flow through a leak for both gasses. Theoretically it can be calculated that the ratio would be close to three. This factor could be used to identify if existing valves can be used, with respect to leakages, in a hydrogen distribution network. In this research there were insufficient leakages measured to confirm or disprove this ratio. For the ten valves at which measurements were done, it can be concluded that if no leak can be measured with natural gas, also no leak can be measured with hydrogen. Because of the limited number of measurements and the absence of leakages, this cannot be guaranteed for the entire population of valves. It is therefore recommended that if a pipeline will be assigned to transport hydrogen, all the valves connected to this pipeline should be inspected individually according to the measuring method described in this report.

From 37.000 installed valves in the national gas grid with a pressure rating above 16 bar, a selection of 12 ball valves and 4 plug valves was made to represent the existing valves as good as possible. The selection criteria are based on the amount, and the most used brands/types installed in the grid. The measurements are divided between external leakages (from the valve to the environment) and internal leakages (from one side of the valve to the other). When a significant external leakage is detected using the “Leak Detection And Repair” (LDAR) method, an additional measurement is executed with a “Hi-flow sampler” (HFS) and/or a flow measurement. Two ball valves are found to have an internal leakage that was of such an order that the valve could not be filled with hydrogen. Therefore, out of the 12 ball valves, there were 10 measurements with hydrogen. The measurement of one of the selected plug valves could not be done, and four plug valves were additionally measured. This makes that in total 19 valves have been used in the measurements.

Plug valves are only tested on external leakages with natural gas. For plug valves it was unfeasible to perform any type of measurement with hydrogen. A comparison between hydrogen and natural gas could therefore not be made for these types of valves.

Summarizing the measurements:

• Three of the twelve ball valves have a detectable amount of natural gas in the valve stem, all without a measurable flow.
• Four out of ten ball valves have a detectable amount of hydrogen in the valve stem, all without a measurable flow.
• For all ball valves an external leakage could not be detected for both gasses. A ratio of the leakage could therefore not be determined.
• Five out the twelve ball valves had a measurable internal leakage for natural gas, of which three were above the Gasunie rejection limit. The internal leak tests could not be performed with hydrogen.
• Two out of the seven plug valves had a measurable external leakage in natural gas. These tests could not be performed with hydrogen.

Besides the leakage measurements, other properties of the existing valves have been examined. A literature research indicated that no problems are expected concerning the hydrogen suitability of the materials used in valves, given the operating pressure and temperature. Manufacturers where inquired about the hydrogen capability of valves. The general response was that no problems are expected when hydrogen is used in the newer valves