SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Ray V. Steele
Organization:Ebara Technologies, Inc.
Date posted: Thu Aug 24 16:47:30 US/Eastern 2000
Subject: Equipment Reliability Database
I have been asked to develop a database to monitor many pieces of equipment at our facility. These would include a number of air conditioners, compressors, cnc machinery, etc. All of this equipment has been online for a number of years. What suggestions might anyone have with regard to types of useful data to collect? Can MTBF and MTTR be calculated on this equipment? If not, what would be useful? Thanks in advance for your help - Ray
Subject: Data Collection
Reply Posted by: Robert Arno
Date Posted: Tue Aug 29 9:14:59 US/Eastern 2000
There are several methods in collecting useful data. My suggestion is that you collect the basic numbers and develop a database system to calculate the reliability numeric. For reliability, collect the operational hours and the number of failures (define a failure). For maintainability collect the actual repair times to include corrective and preventative maintenance. These numeric will provide you with the basic information to calculate the MTTR and MTBF you are looking for. You also have the option of collecting data based on calendar hours. Most equipment will be impossible to collect operational hours to support reliability calculations. The power industry, as well as the HVAC industry, tends to collect the data in calendar hours. This provides an availability of the equipment based on operational and non-operational periods. There are some inherent problems in calculating availability this way but in general it is a sound approach.
Remember that on equipment with inherently long lives you will be collecting a WINDOW of data. Don't be discouraged if the initial reliability numeric for this equipment does not appear reasonable, just keep collecting data. Data collection and analysis is a statistics game, use the basic concepts first to get you in the ball park and then get fancy trying to determine specific problem sources.
Meaningful data collection takes time!!! Use the RAC's Tool Kit to help you with the formulas to set up the database calculations.
I have been in the reliability field for 24 years and have developed several data collection systems. Currently I am completing a database development for Power Distribution, Power Generation, and HVAC equipment on an Access database for accessibility. You can read about it in the IEEE, IAS, conference notes of May 2000.
Subject: Re: Equipment Reliability Database
Reply Posted by: Marina Karyagina
Date Posted: Wed Sep 13 22:56:22 US/Eastern 2000
One usually first decides on the objective of data collection - then you would know what information to collect.
if you are interested in improving reliability of equipment at your facility with the objective of reducing cost of maintenance have a look at some off-the-shelf computerised maintenance management systems (CMMS). There are some quite sophisticated, easy to use and inexpensive packages available.
Information usually stored/collected by these packages are:
-equipment register (the list of your equipment with equipment numbers and locations, technical data ect.);
-routine maintenance actions for each piece of equipment;
-equipment failures and repair actions (usually captured through work orders)
and much more;
Failures can be classified in different categories (electrical, mechanical etc.), failure causes can be recorded and repair actions taken to restore the equipment.
Main information collected for repair actions and preventative maintenance actions are materials used, labour hours spent, total downtime etc. so that one can calculate maintenance costs and cost of downtime.
You might want to collect hours run for each equipment, but if equipment is used pretty regularly then calender hours will do.
my personal view is that looking into the root causes of significant failures (that is failures which have high repair/downtime costs) is more important than compiling MTBFs/MTTRs.