What a dry gas seal application in 1986 showed

Published on: 

The Kårstø K-Lab dry gas seal in Norway was among the first Dry Gas Seal installations with a high sealing pressure as 132 Bara. In 1986, when the selection was made, there were only 8 dry gas seals applications above 80 bara, and none over 110 bara. The seal manufacturer proposed a triple seal for this application.  At the time, a single seal could only accept up to approximately 80 bar pressure differential. The main pressure limiting factors were the distortion of the primary seal rings, pressure distribution in the seal gap and O-ring material for secondary seals. The sealing pressure was broken down to atmospheric pressure over two seals, with an intermediate pressure of 76 bara, which was controlled by an external pressure control valve. The compressor with triple dry gas seals went into operation in Kårstø test loop (K-Lab) for pipeline flowmeter calibration in 1988.

This article contains excerpts from an ASME/Turbo Expo 2019 paper, “Offshore experience with thermal turbomachinery in Norwegian petroleum industry through 50 years. A review of major milestones, lessons learned and impacts,” by Arne Lynghjem, Lars E Bakken and Tore Naess.

The dry gas seals had been selected mainly because it could eliminate oil leakage into the test loop, which was favorable for the calibration process. The compressor was driven by a VSD (Variable Speed Drive) motor, which was the first VSD installation in Norway. The compressor main data is given in Table 5. A back-up oil film seal and seal oil console was provided for risk-reduction purposes, but was never in use. However, it made it possible to do a realistic comparison between the two sealing systems. It was not available any international standards for dry gas seal systems in 1988, and the system design represented a pilot project with participation of seal manufacturer, compressor manufacturer and operator.

The API 614 standard addressed dry gas seal designs for the first time in 1999.  In 1990s and later, the understanding of seal function and introduction of new materials have seen tremendous progress. After year 2000, the seal manufacturer could therefore introduce new dry gas seals with improved materials that managed to handle pressures up to around 400 bar over a single stage.


For an offshore operator, it is essential to fulfil requirements to safe and reliable operation, low weight, small footprints, low leakage rate, and low power consumption. Comparison between the K-lab dry gas seals and the back-up oil film seal system demonstrated substantial savings in favor of dry gas seals:          a) Weight savings: 25.5 %.     b) Space / area savings: 33 %.     c) Running costs savings: 70 NOK/h (estimated).     d) The power consumption of dry gas seals was tested in workshop and it was very low, approximately 6 kW or 0.2% of driver rating. There are not any external power consumptions (in comparison to seal oil system’s need for pumps).

e) The leakage rate of dry gas seals represents a 10-fold reduction compared to oil film seals, according to the manufacturer’s estimates. The actual leakage rates were monitored during testing in manufacturer’s factory and later in continuous operation in the field. When monitoring leakages rates in the field it is important to be aware of the effects of the various seal parameters. The seal gap (h) is seen to have the greatest influence on the leakage, and the gap is dependent on speed and thermal distortion. The leakage  rate increase proportional to speed. The very thin running gap between the dry seal faces is normally 2-3 µm. The actual seal leakage rates measured in the K-Lab did satisfactorily concur with the factory test measurements and the manufacturer’s estimates. However, the field measurements were somewhat higher on pressures above approximately 80 bara. This was believed to be related to O-rings or secondary paths leakages.

f) The life of dry gas seals is difficult to quantify. Cleanness and absence of impurities are decisive factors. The secondary seal Orings can be a limited factor. The seal manufacturer recommended removal for maintenance after approximately 3 years of operation.