Most people don’t really understand risk.  Nor do they realize they are flirting with disaster in many day-to-day situations.  As an example, what would be the loss to you and your company if your primary server should overheat and suffer damage?

There are lots of ways to deal with risk, but first you need to understand that risk is comprised of two components:

• The likelihood of an event (this is the probability of something happening)

• The resulting outcome (what will be the effect once the cause triggers)

 The March 2006 article noted that most people are pretty familiar with outcomes.  They can visualize what would happen in the situation where the company server goes down due to catastrophic failure.  Thousands (or millions) of dollars can be lost, along with significant effects on company operations (such as email, databases, accounts receivable and accounts payable, payroll, etc.).  This article expands the discussion of outcomes for the minority who have difficulties in this area.

 Sometimes, it is easy to understand the effect that will ensue if a certain event happens.  For example, if your server overheats it is pretty easy to understand that it will probably stop working.  However, putting a dollar figure on this event is a bit more difficult to define.

 An anecdotal example from the nuclear power industry illustrates the point very well.  Nuclear power plants require refueling every 18 months.  The refueling processes, called "outages," are scheduled during the Spring and Fall when the demand for power is lowest.  Some plants can be refueled in three weeks, while others take six weeks or longer.  It depends on the type of reactor, the type of fuel used, and what other modifications or upgrades are taking place at the same time.  Since these outages are planned well in advance, the plant can also plan for maintaining the other parts of the generating system, including steam generators, turbines, reactor vessel heads, etc.  These components are not replaced as often as fuel, but still need to be replaced every so often.  Sometimes they have approached the end of their useful life, while other times it is more economical to bring in a new, higher technology (more efficient) unit to replace a less efficient one.  There is also a significant amount of inspection and repair that takes place during an outage.  When maintenance teams find any anomalies, they must be prepared to repair or replace the designated component. 

Left Photo: Cooling tower, turbine buildings, and reactor vessels of a typical nuclear power plant.
Right Photo:  Comanche Peak Nuclear Plant (shown under construction)

When outages run longer than expected, it costs the company in terms of downtime, lost generating capacity, lost revenues/profits, as well as the unanticipated cost of repairs.  Everyone dealing with an outage is very aware of what it costs the company when problems are encountered that affect downtime.  Yet even so, they sometimes make the wrong choice.

During an outage, a team was doing inspections on the steam generators.  They found that the bearings on one generator were wearing out faster than the bearings on the other generators.  The difference was small, but there was still an anomaly.  The estimated cost to disassemble the generator, test it to discover the source of the problem, make any needed repairs, and put the generator back into service was approximately $100,000.  Plant management felt that although there was a slight probability that the generator would fail, they felt the expense was excessive and wanted to get the plant back online quickly.  They chose not to find out why the bearings were showing extra wear, and decided instead to monitor that generator closely and compare the rate of bearing wear with the rates from the other generators.

When the bearings suddenly failed and the steam generator suddenly became useless hulk of twisted metal, the one time expenditure of $100,000 was recognized as a very small price to pay to ensure the proper operation of the generator.  The loss of the generator caused an unplanned outage at the plant.  The generator replacement costs were in the millions of dollars, which does not take into account the millions lost due to the plant being down.

When the consequences are disproportionately high, even very unlikely events can cause huge problems, and need to be reevaluated accordingly.

At the beginning of the article, the question was asked about the loss to the company if the primary server should overheat and be damaged or destroyed.  The costs are going to be higher than most people think.

Everyone can readily understand these costs:

• Replacement cost (to buy a new server)

• Man hours required to physically replace the damaged server

These costs are not so easily understood:

• Administrative costs (finding the right server, placing the order, receiving the server at shipping, getting the server to the techs to install it)

• Downtime due to reprogramming

• Downtime due to loss of data and use of backup data

• Downtime while the server is not in operation (internal - employees and processes)

• Downtime while the server is not in operation (external - customers and vendors)

• Opportunity cost (what "could" have been done with the money instead of replacing the server)

In summary, a small expenditure that protects you from the potential of a catastrophic loss can turn out to be very economical.

Click here for information on this month's product special.  The SH-2 sensor hub is an economical way to ensure your high dollar hardware can be monitored in real time.  Don't put your hardware in danger!  Call  ISC (home of the Rackmount Ranger) at 800-458-6255 for help in selecting the correct product for your application.