Facility Maintenance Rate

Units:	%
Mode:	Input/Output
Multi-band:	True
Default Value:	0
Validation Rule:	Between 0 And 100
Key Property:	No
Description:	Expected proportion of time the facility is unavailable due to maintenance
Detail:

Input

The simulator automatically distributes maintenance events throughout the simulation horizon during the PASA simulation phase. Maintenance Rate is the fraction of time annually that the units are expected to be out-of-service (OOS) due to scheduled maintenance events.

Table 1: Maintenance Rate Example
Property	Value	Units	Band
Maintenance Rate	7	%	1
Min Time to Repair	3	hrs	1
Mean Time to Repair	36	hrs	1
Max Time to Repair	100	hrs	1
Maintenance Frequency	1	-	2
Mean Time to Repair	300	hrs	2

Here the Maintenance Rate = 7%, which implies that on average units will be out of service (OOS) 0.07 x 8760 = 613 hours per annum. The repair time distribution is triangular thus the duration of each outage will vary but on average the unit will be out of service 7% each year.

Note that:

The total discrete maintenance (as defined by Units Out) is summed up and subtracted from the distributed maintenance requirement (defined by Maintenance Rate).
Maintenance events are distributed according to a regional Maintenance Factor set by PASA.
The outage duration function used is set by the properties described in the article Planned and Random Outages.
If Mean Time to Repair is not defined in multi-band case, it will take the value from previous band.

Different types of maintenance can be modeled on a Facility by separating the outage types using different band numbers as in the example above. In band two there is a 'major survey' of 300 hours duration that must be performed once per year: set by the Maintenance Frequency. In general however, it's recommended to set the Maintenance Frequency in band 1, this way we do the major survey first, which could avoid a situation that after maintenance events scheduled by Maintenance Rate, the remaining time space becomes fractional that may not be enough to allocate a major maintenance.

Note that defining Units Out and Maintenance Rate in the same band or different two bands will result in different maintenance requirements in PASA formulation. Here is an example:

Table 2: Units Out and Maintenance Rate Example
Property	Value	Data File	Units	Band	Date From	Date To	Timeslice	Scenario
Units	1		-	1
Max Capacity	500		-	1
Units Out	1		-	1	2021-06-01 00:00:00	2021-06-29 00:00:00
Maintenance Rate	8		%	1				MB1
Maintenance Rate	8		%	2				MB2
Forced Outage Rate	3		%	1				MB2
Forced Outage Rate	3		%	2				MB1

With Units Out and Maintenance Rate defined in the same band (i.e. Scenario "MB1"), the total required maintenance is calculated as:

Maintenance Rate x Capacity x Total Hours = 0.08 *500 *8760 = 350400

This includes discrete maintenance (defined by Units Out) = 500 * 696 = 348000 and distributed Maintenance (the remaining 350400 - 348000) =2400 .

When Units Out and Maintenance Rate are defined in different bands (i.e. Scenario "MB2"), the total required maintenance is calculated as:

Maintenance Rate x Capacity x Total Hours + Total Out Capacity = 350400 + 348000 MWh = 698400

This includes discrete maintenance 348000 and distributed Maintenance 350400 .

Note:

The PASA phase runs across the entire simulation horizon in annual steps. If the horizon is less than one year there will be some distortion in the placement of distributed maintenance events, since they will not be distributed with reference to the entire year.

The automatic scheduling of distributed maintenance can be switched on/off using the Outage Scope of the Stochastic objects associated with the Model being run. Discrete maintenance events (defined by Units Out) can only be disabled by either removing the Units Out properties or tagging them with a Scenario and omitting that scenario from the Model being run.

Output

As an output, Maintenance Rate is the percentage of time the unit is out-of-service on maintenance either as a result of defining Maintenance Rate or Units Out.