الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 94 |  |  |
| | | from | 94 | | |
Abstract
Fuzzy Layer of Protection Analysis Applied To Fuel Gas Systems
(September 2011)
Mohamed Khalil Ali, B.S., Alexandria University
Natural gas is a clean and cheap source of energy that has a great perspective for development in the future. Natural gas is processed in special processing plants to produce LPG and methane gas. These natural gas plants demand high amount of energy which is provided through dedicated fuel gas units inside the plant. The importance of the fuel gas unit is that it’s the heart of the plant without which the whole plant will shutdown the massive loss in production and money in such case has increased the need for a thorough risk assessment for the associated hazards in the unit operation and design.
Managing risk requires certain safety functions in order to prevent and/or mitigate accidents and undesired consequences. These safety functions have to be at a certain reliability and integrity level (SIL) to bring the risk to the acceptable levels. The level of integrity is reflected on the amount of effort exerted on the design, maintenance and operation of the safety function which eventually controls the cost of the safety function.
There are several methods to determine the SIL rating of SIF which ranges from qualitative methods like risk matrix and risk graph to quantitative methods such as FMEA passing through semi quantitative ¬semi qualitative methods like LOP A.
LOP A is an efficient technique to determine the associated risks with industrial activates and the level of protection needed to bring this risk to the acceptable level however the lack of accurate and definite information limits the effectiveness of LOP A and any other risk assessment method and since the traditional thinking states that uncertainty is equivalent to risk this leads to more conservative safety measures and subsequently additional risk.
The development of fuzzy logic was a breakthrough in mimicking the way human think for use in computers, it provided logic developers and programmers with a mean to use linguistic variables and uncertain data in