Predictive Modeling in Fugitive Emissions Testing
Abstract
Greenhouse gases have become an increasingly significant issue in the last few decades. As a response, many organizations have sought to tighten their regulations on their operations to reduce their contributions to greenhouse gases. The International Organization for Standardization has a standard 15848-1 that classifies industrial valves for the oil and gas industry. They too have aimed to tighten their regulations, including this specific standard. However, the current requirements from ISO 15848-1 has made it extremely difficult for manufacturers to get any industrial valves and seals passed. This begs the question, are the new tightness classifications for the ISO standard appropriately relating the test gas, Helium, to the allowable Methane leakage concentration? And with that, is Helium even the best option for a testing gas? To investigate this, a series of experiments were conducted to collect Helium and Argon leak rate data under many temperature and pressure conditions. With this data, a Helium/Argon leak rate ratio model was created with machine learning techniques. Using this model, an Ar/CH₄ multiplier, and diffusion modeling, the ISO 15848-1 tightness classes can be assessed for their accuracy. A disconnect between the ISO 15484-1 Helium and Methane requirements has been identified and there is a call to reconsider the Helium requirements. In addition, a suggestion to investigate Argon as an alternative leakage test gas is also raised.
Keywords: fugitive emissions, methane leak testing, ISO 15848
Collections
- OU - Theses [2115]