case study: Refrigeration
Danfoss is a world leader in instrumentation and control systems
for Refrigeration. Additionally it is a major service provider to
the Retail industry, providing management of maintenance requirements
for in store refrigeration units. On average a supermarket has about
100 refrigerated display cabinets and each of these is supported
by controls that monitor temperature and operation of the cooling
units. A service failure in a refrigerated display is a serious
event. Food may wasted and the loss of effective shelf space carries
a significant opportunity cost. One supermarket chain has estimated
that, at any point in time, the total of units out of services equates
to two entire stores.
Danfoss monitors alarms produced automatically by the controls
when temperatures exceed operating parameters and despatches service
engineers to address the problem. Detailed analysis of its’
operations showed Danfoss that more than 60% of all engineer despatches
were in respect of “non critical” alarms. At a cost
of call out of £100 this added up to a highly significant
cost that could be reduced. Further analysis proved that in respect
of “critical” alarms, on average five engineer despatches
were need to resolve the problem. There was obviously a major opportunity
for improving customer service and reducing costs if the alarms
could be predicted and diagnosed before they occurred.
Diagnostic work at the Danfoss Nordborg Research Centre demonstrated
that failure mode can be determined by detailed analyses of the
measured parameters on the control units. Fundamental mathematical
transformations and analyses are run on the data sets and fault
conditions identified. The system is self-learning as it ‘groups’
types of fault mode together and then relates this to the recorded
failure information. From this a matrix of statistical analyses
can be undertaken to provide a rule set for predictive analyses.
These are then used to determine when a system shows the first sign
of degradation as well as identifying the failure mode when failure
occurs. This moves the system away from reactive alarms to pro-active
diagnostic systems to identify the onset of failure and allow the
engineer to schedule the correct maintenance procedure to rectify
the problem. The major benefits being:
- More efficient service
- Most faults identified before failure
- Most faults fixed before disruption
- Fewer emergency call outs
- Visits at scheduled times to suit client
- Fewer wasted visits due to miss-diagnosis
- Fewer return visits
- Predicted mean time between failure
- Reduced time on site to fix problems
Danfoss chose to work with C3 to develop the SmartXplore application,
based upon APM analytics to collect data from control units and
interpret this into information that supported the engineers’
decisions regarding the cause of the alarm and appropriate action.
Danfoss has been able to achieve:
- 70% reduction in call outs
- Increased first time fix rate
- Accurate figures for mean time to failure on components
- Key Performance Indicators on all aspects of refrigeration
- Increased market share because:
Danfoss customers have been able to achieve:
- Increased serviceability on refrigeration units
- Reductions is product losses due to unit failure
- Reductions in power consumption
- Reductions in total cost of ownership.
|
