Risk assessment based on exposure time. 1

[MaintGuy]

I would like to determine a safety factor for equipment failures that would be valuable in making decisions. One very important aspect, is the exposure time. My example, is electrical equipment failure that is in a remote area. Operators walk by the area say once/day for 10 min. How can I quantify the risk involved or probability so that it makes sense. Does anyone have any suggestions or recommended rudimantary calculations? My appologies for being so general.

THanks

 

[aspearin1]

You need to know the probability that the equipment will fail, (in an unsafe manner), either based on experience or manufacturer certifications. You can express this as a % of up-time. You can then find out the probability of someone being in the area, as a % of up-time. If you multiply these together, you get the likelihood that someone will be in the area during the time of unsafe failure.

CAUTION!!!! Safety and statistics don't always mix well. Unfortunately, in my experience, safety professionals don't convert safety risks in $$$ amounts, at least there is no standard metric for this (although insurance underwriters probably do this very well). This means that there is no way to calculate an ROI for safety fixes. This is like Civil Engineers planning for the 100-yr flood....If it never comes, you overdesigned and wasted time and money. But if it DOES come, you'll be glad you paid for it up front. Class action suit may not be so predictable. My suggestion is if there is evidence of a failure waiting to occur, Poke Yoke it.

Aaron A. Spearin
ASQ CSSBB
Engineering Six-S'$

http://www.Engineering6ss.com

"The only constant in life is change." -Bruce Lee

 

[MaintGuy]

Thanks for your response. It certainly makes sense and I will take your words of caution.

 

[PSE]

Good advice from aspearin1. I would add that a potential way to provide a $$$ amount would be to consider the potential fine amounts and medical costs that could be levied or incurred in the event of an accident. As with the 100 year flood example, it might often take just 1 event to make the precautions worthwhile.

Regards,

 

[tab1999]

Larger corporations use Recordable Incident Rates and Lost time incident rates to determine saving, based on avoiding accidents and lost times at given, known associated costs.

I'm thinking that there is a predetermined cost associated with various incidents posted on OSHA.org or check the numbers within your organization.

Avoidance = Value Added

" There is no Shortcut to anyplace worth going "

 

[SJones]

http://www.hse.gov.uk/research/rrhtm/rr025.htm

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdoweb/

 

[tomwalz]

Source: Book ‘Freakonomics’ by Levitt & Dubner

pp. 130 & 131
Chart showing how the state of Connecticut compensates for work related injuries in terms of weeks of pay.

Assumption: Assume an event that happens at any time will injure an employee and the event will be equally dangerous for the full ten minutes.

1. Odds of Occurrence
Machine A has a dangerous failure once a year so 1/144 x 1/365 = 0.0019% or every 52,365 years
Machine B has a dangerous failure once a month so 1/144 x 1/30 = 0.023% or every 4,347 years
Machine C has a failure once a day so 1/144 or 0.7% or once every 142 years

2. Cost of Occurrence
If the accident causes the employee to lose the use of their master hand then, in Connecticut, they will get 168 weeks of pay. Call this $150,000.

Machine A will cost $ 2.85 per year in employee injury
Machine B will cost $ 34.50 per year in employee injury
Machine C will cost $1050.00 per year in employee injury

3. In Washington State we have no fault employee compensation. If I buy a cheap machine and employees get hurt because of it then they can’t sue me and the state will pay them. If I have too much cost then my rates go up but not by nearly enough to cover the cost of treatment. So figure that having an employee injured and losing the use of the master hand will raise my rates by maybe $2,000 a year for a few years.

Machine A will cost $0.04
Machine B will cost $0.46
Machine C will cost $14.00

4. Add machine D which has a failure every hour. 4 times a day an employee will be injured. Except management will instruct employees to wait for the failure and the correct it on their walk by so no one will be injured. (Actually employees will probably develop this in self defense)

5. Randomize the occurrences so employees can’t anticipate them then a new factor enters. If too many employees are injured then the state will eventually take legal action. Assume you can stall for four years for $50,000 per year. In year five you have to go to court which costs S200,000. You are fined $5,000,000 which makes the state look good in the papers. On appeal, four more years at $50,000 per year this is reduced to $500,000.

This is $600,000 in legal fees over 9 years plus a $500,000 fine paid in year ten for a total of $1,100,000 over ten years. This is $110,000 which still makes it cheaper than paying for an injured employee.

6. If the employee is injured on a walk-by then the amount of injury will be based on proximity. You need a bell curve with a big dip in the middles, sort of a camel curve. In minutes 1 and 10 the employee is mildly injured. At minute five the employee is killed. Generally killing an employee, especially in jury trials, is cheaper than severely injuring an employee. (You can’t bring a corpse to court but you can plant a wheel chair front and center.)

7. In real life you don’t worry about employee injury in this instance since the chances are so small. In college I worked in a food plant made of concrete block. Across from the pressure retorts they had a wall made of 2x4’s and fiberglass panels. If the retort doors blew off they would go through the walls and cause maybe $200 of easily repaired damage. There was a corridor between the retort door and the wall but there was no effort to protect the employees in case the doors blew off. The walls would always be there but people were very seldom there.

8. Killing an employee can still be good business as witnessed by the recent coal mine collapses. Killing employees slowly is still good business as evidenced by most of the coal industry.

9. Never talk about an acceptable level of danger. You will find yourself on a witness stand and the attorney will ask you if you knowingly bought machine knowing that it would eventually kill your employees. You will not be able to explain that statistically the odds were only once ever 52,000 years.

The Good News
Machines that fail badly enough to injure employees typically cause damage to themselves and seriously interfere with production schedules. (E.G. this is a press that shatters a die and blows shrapnel at high velocity. This is different than an unguarded press that can crush hands as part of its normal operation.)

Fifteen years ago I decide to address the dangers from coolants exposed to carbide. (Hardmetal Disease, a nasty emphysema like death, caused by cobalt exposure caused by cobalt leaching.) Simple coolant filter units could save a lot of lives so I built some and actually sold a few based on safety and health. That’s all I could sell and my only distributor was afraid even to discuss safety and health any more because of potential liability.

We switched and we now sell based on economic advantages in terms of consumables, rebuild, process times, etc. As an example, today we sold 25 units to a man who was concerned with cleanliness and health when he called a few months ago. After trying a few things we had a new design for him and he could go to purchasing and talk about diamond wheel life, coolant life, increased processing speed, machine maintenance and downtime. He didn’t have to mention safety and health once to justify the purchase.

The point is that addressing safety and health exposes you to considerable liability because you can never make things completely safe and the juries never understand statistics. What you should do is design very efficient, cost effective machines and processes and sneak in employee safety and environmental concerns but never mention them directly.

I would suggest you consider designing a very cost effective solution on paper for management and accounting while also and making sure it saves lives and creates a better world for yourself. As near as I can tell there are a bunch of folks here doing this. Once you get in the mindset it’s not all that hard, it saves the company a lot of money, it is personally satisfying and, just maybe, it really is truly fine engineering.


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com