Hydrogen tightness can be calculated!

Technical tightness of flange connections according to DIN EN 1092-1 to 4
Determination of hydrogen leakage based on DIN EN 13555
Calculation according to DIN EN 1591-1

The energy transition in Germany is leading to the construction of numerous hydrogen production plants. Hydrogen produced from sustainable energy sources is considered a foundation, either to balance fluctuations in energy generation or to serve as a raw material for additional processes. The centerpiece of these plants is always the electrolyzer, which transports the generated hydrogen through piping to the other components of the facility. Typically, the builders of these facilities face the challenge of providing evidence of tightness or demonstrating the absence of leaks for the entire plant.

Definition of technical tightness:

A flange connection according to DIN EN 1092 is considered (permanently) technically tight according to the current state of technology if a computational proof according to EN 1591-1 or Finite Element Analysis (FEM) for a leakage class L0.01 can be provided (TA-Luft edition August 18, 2021, VDI 2290 edition June 2012). This generally applies to H2 applications as well. However, the sealing characteristics underlying the flange calculation are usually determined according to DIN EN 13555 with helium. Helium is the chemical element with the second-lowest density after hydrogen and is closest in size to hydrogen (see Figure 1). Due to the different values for gas viscosity and gas density with approximately the same kinetic diameters, different leakage rates can be expected in individual cases.

A universally applicable conversion factor compared to helium cannot be provided based on current knowledge. This factor depends on many additional parameters, such as flow conditions, permeation and adsorption processes in the sealing material, etc..

Under these conditions, what options are available for verification of flanges in accordance with DIN EN 1092-1 to -4?

Option 1: The type-based component test to determine the leakage rate of a flange connection. For the test situation, the achievable surface pressure of the weakest flange in the PN series is assumed. This surface pressure, Qmin(LComponent test), serves to demonstrate in a laboratory test using a mass spectrometer with the test medium H2 under system pressure that the combination of flange, gasket, and bolts complies with the required tightness class under the given process conditions. These tightness classes can be either according to DIN 3535-6 L0.1 with a specific leakage rate ≤ 0.1 [mg s−1 m−1] or TA-Luft L0.01 with a specific leakage rate ≤ 0.01 [mg s−1 m−1], or higher, see DIN EN 13555 Table 1 tightness classes. Prior to this, temperature aging up to the maximum process temperature must be performed to simulate system relaxation. Based on this evidence, a torque table is to be developed that guarantees a minimum assembly surface pressure Qmin (LTorque table) for all other nominal sizes in the PN series, where: Qmin(LComponent test) ≤ Qmin(LTorque table). Proper and quality-controlled assembly with corresponding documentation is a prerequisite for such evidence. This approach represents an additional effort for the plant builder/operator and is therefore likely not to be used frequently for verification.

Option 2: Individual measurement at the flange in the plant under operating conditions. This option likely represents the most elaborate approach, as it involves measuring the resulting leakages under operating conditions using the purging gas method. It is necessary to enclose the flanges to be measured. This method also entails a significant additional effort for the plant builder/operator and is therefore likely not to be used frequently for verification.

Option 3: Calculation according to EN 1591-1 is the most common method for verification. In the design of the plants, a strength verification of the flange system must be provided. It is advisable to use the tightness class verification through a calculation according to EN 1591-1 or Finite Element Analysis (FEM) for a leakage class LN, as both methods, according to VDI2290, demonstrate both the strength of the flange connection and the tightness. The relevant values for the calculation and the verification of tightness with respect to the leakage class L are the sealing characteristics, such as minimum surface pressure in the assembly condition Qmin(L) and the minimum surface pressure in the operating condition QSmin(L), depending on the initial surface pressure QA.

The sealing characteristics according to EN 13555 were previously available only from measurements with helium. However, it was not certain whether these characteristics could be applied to hydrogen. Therefore, KLINGER Dichtungstechnik decided to conduct tests according to DIN EN 13555 using hydrogen as the test medium, deviating from the recommended test medium helium, for several selected fiber and PTFE materials from the product range. These tests were carried out by the accredited independent testing institute AMTEC Messtechnischer Service GmbH. The aim of the investigations was to provide a reliable comparison of both measurements since these values are relevant for the tightness and strength, and hence the safety, of the flange connection.

Results and Insights

In many cases, a significant correlation was observed between the curves, as seen in the example of KLINGERSIL®C-4430 (see Figure 2). However, there are also measurement values that demonstrate substantial differences between the measurements, as evident in the example of KLINGER®topchem 2003. In this specific case, the hydrogen leakage curve is approximately an order of magnitude below that of helium (see Figure 3).

With the obtained measurement results, KLINGER can perform calculations according to EN 1591-1 for pressure levels of 10 bar and 40 bar using both helium-based and hydrogen-based sealing characteristics, thereby precisely demonstrating the leakage class in conjunction with the flange strength. With all tested sealing materials, it is possible to comply with the requirements of TA-Luft and, consequently, DIN 3535-6.

Author

Dipl. Ing. Stefan Keck

Product Manager Seals (Hbv.) Sealing Technology,                                                                                                                                                                  KLINGER Germany

Picture 4

Sources:
TA-Luft edition August 18, 2021,
VDI 2290 June 2012 edition
www.arnold-chemie.de
DIN 3535-6
DIN EN 13555:2021

Picture directory
Picture 1: Table of kinetic diameters of molecules and noble gases ©www.arnold-chemie.de
Picture 2: Leakage curves KLINGERSIL® C-4430 ©KLINGER GmbH
Picture 3: Leakage curves KLINGER®top-chem 2003 ©KLINGER GmbH
Picture 4: Stefan Keck ©KLINGER GmbH