Saturday, November 3, 2012

Availability Makes a Difference … (Part II)









The Modular Airborne Fire Fighting System II (MAFFS II) being installed in a C-130 at Peterson Air Force Base, CO in 2009.[xxii]  (Photo courtesy of 302 AW, Air Force Reserve Command)
  
The MAFFS II as installed in a C-130 cargo bay, Peterson Air Force Base, CO in 2012.[xxiii]  (Photo courtesy of 302 AW, Air Force Reserve Command)
On 23 June 2012, a wildfire was quickly growing in the vicinity of Colorado Springs, CO.  Ultimately known as the Waldo Canyon fire, this fire lead to the U.S. Forest Service’s (USFS) 24 June 2012 decision to activate its Modular Airborne Fire Fighting System II (MAFFS II) for the 2012 wildfire season.[i] “On June 25 … MAFFS-equipped C-130s … received their first launch orders for 2012 and began to fly fire suppression missions on what would become the costliest wildland fire in Colorado’s history.”[ii] By 17 September 2012, when the MAFFS was deactivated for 2012, eight MAFFS equipped C-130s had “released almost 2.5 million gallons of fire retardant during 1,011 [aerial] drops on fires in 10 states.”[iii]
So, what is the MAFFS system? 
“MAFFS is a joint Defense Department and U.S. Forest Service program designed to provide additional aerial firefighting resources when commercial and private air tankers are no longer able to meet the Forest Service's needs ….
As a self-contained aerial firefighting system owned by the U.S. Forest Service, MAFFS can discharge 3,000 gallons of water or fire retardant in less than five seconds, covering an area a quarter of a mile long by 100 feet wide. Once the load is discharged, it can be refilled in less than 12 minutes.”[iv]
There are three Air National Guard wings and one Air Force Reserve Command wing that are MAFFS-qualified and provide the C-130 aircraft when the USFS activates MAFFS.[v] 
Why are these aerial firefighting systems so critical? According to the USFS’s January 2012 Large Airtanker Modernization Strategy, “Airtankers are used to deliver fire retardant to wildfires, thereby reducing fire intensity and rate of spread until ground personnel can reach the fire. Airtankers play a key role in successful initial attack, which is one of the most difficult and critical components of wildfire management.”[vi] The USFS report goes on to state that up to 70,000 communities across the U.S. are at risk of wildfires, and the annual costs of suppressing and recovering from wildfires “amounts to billions of dollars each year.”[vii]
Like any complex mechanical system, maintaining the MAFFS units and keeping them available requires skilled technicians, proper tools and equipment, and a ready source of spare parts. A 7 July 2012 article in The Seattle Times notes that the “Forest Service has stockpiled enough major parts, can source many smaller parts, and can mend the biggest parts no longer being made to keep the system running”[viii] according to Scott Fisher, MAFFS coordinator for the Forest Service. The article continues, quoting Mr. Fisher, "The system was built for at least 20 years … I would not be surprised to see this thing fly for a full 30 years."[ix] 
Of course, to make that projection a reality is where applying readiness-based sparing (RBS) and its availability-based sparing recommendations would be invaluable. For example, RBS can determine the proper range and depth of spare parts to keep the MAFFS units fully operational and available to continue their wildfire suppression mission for another 20 to 30 years.

System availability revisited

As noted in a previous posting on this blog, the availability of a fielded system in normal operations is dependent upon its supporting maintenance infrastructure. There are two basic categories of maintenance. The first category, corrective (or unscheduled) maintenance, generates independent part demands which must be forecasted. The second maintenance category, preventive (or scheduled) maintenance, generates dependent parts demands whose timing and quantities are derived from the maintenance schedule. Modern RBS applications can accommodate both of these maintenance types when sizing supply support.  Let’s now take a closer look at some well-known availability relationships.

Inherent availability

When it is important to examine the reliability-maintainability trade-offs for a system’s design, then inherent availability is an appropriate system performance measure.[x] Inherent availability [Ai] is calculated as (EQ. 1)  [xi] 
 












where the mean time between failures (MTBF) is (EQ. 2) [xii]


and mean corrective maintenance time  (also called Mean Time to Repair) is “For a sample of repair actions, a composite value representing the arithmetic average of the maintenance cycle times for the individual actions.”[xiii]

Operational availability

Unfortunately for operators and logisticians, the systems they support don’t operate in such an ideal environment and operational availability [Ao] becomes the more appropriate measure (EQ. 3) [xiv]



 











where MTBM is the mean time between maintenance. According to the DoD Reliability, Availability and Maintainability (RAM) Guide, MTBM considers “all maintenance actions, including repairing design/manufacturing failures and maintenance-induced failures, performing preventive maintenance, and other actions (e.g., remove an item to facilitate other maintenance).” [xv]
Mean maintenance downtime (MDT) measures “the total elapsed time required (when the system is not operational) to repair and restore a system to full operating status, and/or to retain a system in that condition. MDT includes mean active maintenance time (M bar), logistics delay time (LDT), and administrative delay time (ADT).[xvi]  In this definition, M bar is the actual hands-on maintenance time [xvii]; LDT is the time maintenance is delayed due to the lack of a spare part, maintenance facilities or equipment; or transportation [xviii]; and ADT is the time maintenance is delayed due to personnel availability, workload scheduling, etc.[xix] It is important to note that implicit to these availability calculations is the assumption that they represent the long-run average performance for a system.

Availability applications

The above measures describe availability from a traditional logistics viewpoint. RBS practitioners have adapted these definitions to support their day-to-day management of reparable item inventory systems, viewing availability as the product of maintenance and supply availability. In later posts we’ll define supply availability from the RBS perspective more precisely, but for the time being a reasonable working definition is “What is the probability that any randomly selected end-item (e.g., an aircraft) in a fleet is not down due to the lack of a spare part.” 
Stockage policy has a dramatic effect upon supply performance, and is how RBS influences the systemwide availability performance measure. Accurately estimating this performance is critical, especially when you are sparing for a small fleet of high demand assets such as the MAFFS II units and their C-130 aircraft. If the fleet has insufficient availability, then missions may be scrubbed and lives and property (in the case of wildfire fighting) will be placed in jeopardy.

Postscript

On 1 July 2012, MAFFS 7, the call sign for a MAFFS II equipped C-130 from the 145th Airlift Wing at Charlotte-Douglass Airport, NC, crashed while performing its fourth airdrop of the day over a South Dakota wildfire.[xx] [xxi] Four North Carolina Air National Guard crewmembers were killed – Lt Col Paul K. Mikeal, Maj Joseph M. McCormick, Maj Ryan S. David, and SMSgt Robert S. Cannon and two crewmembers were seriously injured (names not released). 
For more information about MAFFS 7, there is a very touching memorial page to this brave crew at http://www.nc.ngb.army.mil/PAO/News/Pages/MAFFS.aspx . Our hearts and prayers go out to the members, families and friends of this crew.



[i] Skarban, Ann. “MAFFS 2012,” Citizen Airman, Vol. 64, No. 5: 14-17 (October 2012).  Downloaded from http://www.citamn.afrc.af.mil/shared/media/document/AFD-120927-026.pdf on 31 Oct 2012.
[ii] Skarban (p. 14).
[iii] Air Force News Service (AFNS), “Forest Service deactivates C-130 firefighting operations,” 17 Sep 2012.  Downloaded from http://www.af.mil/news/story_print.asp?id=123318258 on 20 Sep 2012.
[iv] AFNS, 17 Sep 2012.
[v]  U.S. Air Force, “Fact Sheet: Modular Airborne Fire Fighting System,” (undated) downloaded from http://www.af.mil/information/factsheets/factsheet_print.asp?fsID=10566&page=1 on 13 July 2012.
[vi] USDA Forest Service.  Large Airtanker Modernization Strategy.  17 Jan 2012. (p. 2)  Downloaded from http://www.fs.fed.us/fire/aviation/airtanker_modernization_strategy.pdf on 26 Aug 2012.
[vii] USDA Forest Service (p. 3)
[viii] Gruver, Mead, “Some experts worry about a lack of parts for C-130 wildfire sprayers,” The Seattle Times, 7 July 2012.  Downloaded from http://seattletimes.com/html/nationworld/2018631369_forestfires08.html on 5 Sep 2012.
[ix] Gruver, 2012.
[x] Ebeling, Charles E. An Introduction to Reliability and Maintainability Engineering. Long Grove, IL: Waveland Press, Inc., 1997. (p. 255)
[xi] Blanchard, Benjamin S.  Logistics Engineering and Management (Second Edition).  Englewood Cliffs, NJ: Prentice-Hall, Inc., 1981. (p. 66)
[xii] Blanchard (p. 400)
[xiii] Department of Defense. DOD Guide for Achieving Reliability, Availability, and Maintainability. Office of the Under Secretary of Defense (Acquisition, Technology and Logistics).  Washington: 3 August 2005. (p. 3-8)
[xiv] Blanchard (p. 67)
[xv] DoD RAM Guide (p. 3-11)
[xvi] Blanchard (p. 46)
[xvii] Blanchard (p. 45)
[xviii] Blanchard (p. 46)
[xix] Blanchard (p. 46)
[xx] Church, Aaron M., ed. “Four Airmen Die Fighting Western Wildfires,” ­Air Force Magazine, Vol. 95, No. 8: 10 (August 2012).
[xxi] Schogol, Jeff, “Legacies remain for airmen killed in C-130 crash,” Air Force Times, 16 Jul 2012, p. 16.
[xxii] Air Force Reserve Command, downloaded from http://www.302aw.afrc.af.mil/shared/media/photodb/photos/110909-F-XU932-0171.JPG  on 12 Sep 2012. 
[xxiii] Air Force Reserve Command, downloaded from http://www.302aw.afrc.af.mil/shared/media/photodb/photos/2012/07/120627-F-IG195-644.jpg  on 12 Sep 2012. 

  





Sunday, July 1, 2012

Availability Makes a Difference …


The DART II Tsunami System.[i]  (Diagram courtesy of NOAA)
On 11 April 2012 at 2:38:37 p.m. local time (08:38:37 UTC), approximately 269 miles off the west coast of northern Sumatra, Indonesia[ii] there was a magnitude 8.6 Richter Scale earthquake.[iii] Within minutes, buoys from the Deep-ocean Assessment and Reporting of Tsunamis (DART®) system detected and reported ocean level changes resulting from the earthquake, enabling the DART system to accurately predict that the resultant “tsunami would not be big.” [iii] Given this region’s experience with tsunamis, this was vital information that made a difference for thousands of people!
DART buoys are positioned at 54 locations around the world and are operated by the U.S. National Oceanic and Atmospheric Administration (NOAA) and six other countries.[iv] “Originally developed by NOAA, as part of the U.S. National Tsunami Hazard Mitigation Program … the DART® Project was an effort to maintain and improve the capability for the early detection and real-time reporting of tsunamis in the open ocean.”[v] Since 2008, the National Data Buoy Center (NDBC) has operated 40 DART II buoys, primarily around the Pacific Rim and along the U.S. eastern seaboard.
As the diagram above shows, each DART II buoy location “consists of two physical components: a tsunameter on the ocean floor [the bottom pressure reader (BPR)] and a surface buoy with satellite telecommunications capability. The DART II systems have bi-directional communication links and are thus able to send and receive data from the Tsunami Warning Center and others via the Internet.”[vi]
The BPR collects temperature and pressure data which are converted to an estimated sea-surface height every 15 minutes while the system is operating in standard mode reporting.[v] With this data, the “Tsunami Detection Algorithm works by first estimating the amplitudes of the pressure fluctuations within the tsunami frequency band, and then testing these amplitudes against a threshold value…. If the amplitudes exceed the threshold, the tsunameter goes into Event Mode to provide detailed information about the tsunami.”[vii] Once the BPR enters into the event mode “15-second values are transmitted during the initial few minutes, followed by 1-minute averages. Event mode messages also contain the time of the initial occurrence of the event.” [v]
For the 11 April 2012 earthquake, a DART II station in the Bay of Bengal reported the event 2:23 minutes after the earthquake. The next closest DART II station (600 nautical miles West-Northwest of Phuket, Thailand) reported an event after an additional 1:15 minutes. A third DART II station reported an event after another 2:15 minutes. Thus, in less than 6 minutes, three DART II stations had recorded this earthquake and transmitted wave height information to the Tsunami Warning Center enabling the Center to determine that the resultant wave would have limited destructive potential.

Relating “readiness” and “availability”

Although the 11 April 2012 tsunami determination was timely and effective, when the earthquake occurred, only 38 of the 54 DART II buoy locations worldwide were on-line and reporting wave height data — a systemwide point availability of 70 percent. This compares to an 80 percent reliability and data return (i.e., availability) goal for the DART II system.[viii]
The relationship between readiness and availability was examined in an earlier posting on this blog. Readiness was defined as the “ability to provide capabilities required by the combatant commanders to execute their assigned missions. This is derived from the ability of each unit to deliver the outputs for which it was designed.”[ix] From there, the assessment methods that were described expressed system readiness as a function of personnel, materiel, equipment, and training.
Of course, for readiness-based sparing, the most relevant component of these readiness assessment systems is materiel supportability which directly contributes to equipment availability. According to Blanchard, availability is “the measure of the degree a system is in the operable and committable state … when the mission is called for at an unknown random point in time.”[x]

System availability

The availability of a fielded system in normal operations is dependent upon its supporting maintenance infrastructure. As defined in the U.S. Department of Defense (DoD) reliability, availability and maintainability (RAM) guide “Availability as measured by the user is a function of how often failures occur and corrective maintenance is required, how often preventative maintenance is performed, how quickly indicated failures can be isolated and repaired, how quickly preventive maintenance tasks can be performed, and how long logistics support delays contribute to down time.”[xi]

Maintenance categories

There are two basic categories of maintenance. The first category, corrective maintenance, is “the unscheduled actions accomplished, as a result of failure, to restore a system to a specified level of performance.”[xii] Part demands arising from corrective maintenance are referred to as independent demands which must be forecasted. The following figure illustrates the main components of the corrective maintenance cycle at an operational location for a first level indenture line replaceable unit (LRU).[xiii]
The Corrective Maintenance Cycle.
The second maintenance category, preventive maintenance, is “the scheduled actions accomplished to retain a system at a specified level of performance by providing systematic inspection, detection, servicing, condition monitoring, and/or replacement to prevent impending failures.”[xiv] Parts demands arising from preventive maintenance are commonly referred to as dependent demands (since they are determined by the scheduled maintenance activities) and their quantities may be derived from the maintenance schedule.

Availability measures

The support environment predictability that the two categories of maintenance are performed in is the fundamental difference between inherent and operational availability – two of the most common measures of system availability.
When it is important to examine the reliability-maintainability trade-offs for a system’s design, then inherent availability is an appropriate system performance measure.[xv] As defined by Blanchard, inherent availability “is the probability that a system or equipment, when used under stated conditions in an ideal support environment (i.e., readily available tools, spares, maintenance personnel, etc.), will operate satisfactorily at any point in time as required. It excludes preventive or schedule maintenance actions, logistics delay time, and administrative delay time….”[xvi]
Unfortunately for operators and logisticians, the systems they support don’t operate in such an ideal environment. As the DoD RAM Guide notes, operational availability becomes the more appropriate measure when “the effects of design and the support system on availability are being considered….”[xvii]
Blanchard defines operational availability as the “probability that a system or equipment when used under stated conditions in an actual operational environment, will operate satisfactorily when called upon.”[xviii] The availability definition has now expanded to include the impacts of both maintenance categories – corrective and preventive, and the “actual” operating environment is affected by all of today’s logistical realities “the availability of spare and repair parts, tools, support equipment, and maintenance personnel….”[xix] This is the environment where the effectiveness and efficiency of readiness-based sparing comes into play.

Summary

Clearly, for a complex system (such as the DART® system) to make a difference, it must be properly designed, maintained and supported to be available when needed. Furthermore, it’s important to use an appropriate availability measure when designing, building and evaluating this support environment. In the next posting, we’ll take a closer look at how these alternative availability measures are calculated and how they’re routinely employed by today’s logistics practitioners.


[i] NOAA Center for Tsunami Research, DART II System. Downloaded from http://nctr.pmel.noaa.gov/Dart/Jpg/DART_II_metric-page.jpg on 18 Feb 2011.
[ii] U.S. Geological Survey. Magnitude 8.6 - Off the West Coast of Northern Sumatra, Downloaded from http://earthquake.usgs.gov/earthquakes/eqinthenews/2012/usc000905e/#details on 18 April 2012.
[iii] Surveying minimal damage from twin quakes, Indonesians can’t believe their luck,” The Washington Post, 12 April 2012, downloaded from http://www.washingtonpost.com/world/asia_pacific/surverying-minimal-damage-from-twin-quakes-indonesians-cant-believe-their-luck/2012/04/12/gIQAT26UCT_story.html\  on 12 April 2012.
[iv] National Data Buoy Center. Deep-ocean Assessment and Reporting of Tsunami system location map at http://www.ndbc.noaa.gov/dart.shtml accessed on 18 April 2012.
[v] National Data Buoy Center. “Deep-ocean Assessment and Reporting of Tsunamis (DART®) Description.” Downloaded from http://www.ndbc.noaa.gov/dart/dart.shtml on 29 Mar 2011.
[vi] Meinig, Christian., et al.  “Real-Time Deep-Ocean Tsunami Measuring, Monitoring, and Reporting System: The NOAA DART II Description and Disclosure.” 2005.  Downloaded from http://www.ndbc.noaa.gov/dart/dart_ii_description_6_4_05.pdf on 28 Feb 2011.  (p. 5)
[vii] Meinig (p. 8)
[viii] Meinig (p. 6)
[ix] U.S. Department of Defense.  Department of Defense Dictionary of Military and Associated Terms.  Joint Publication 1-02.  Washington: 8 November 2010 (as amended through 15 January 2012).  (Page 277)  Downloaded from http://www.dtic.mil/doctrine/new_pubs/jp1_02.pdf on 18 March 2012.
[x] Blanchard, Benjamin S. Logistics Engineering and Management (Second Edition).  Englewood Cliffs, NJ: Prentice-Hall, Inc., 1981. (p. 20)
[xi] U.S. Department of Defense. DOD Guide for Achieving Reliability, Availability, and Maintainability. Office of the Under Secretary of Defense (Acquisition, Technology and Logistics).  Washington: 3 August 2005. (p. 1-1)
[xii] Blanchard (p. 35)
[xiii] Phillips, Shawn M.  Improving Marine Corps Total Life Cycle Management by Connecting Collected Data and Simulation. MS Thesis. Naval Postgraduate School, Monterey, CA, June 2009. (p. 28)
[xiv] Blanchard (p. 35)
[xv] Ebeling, Charles E. An Introduction to Reliability and Maintainability Engineering. Long Grove, IL: Waveland Press, Inc., 1997. (p. 255)
[xvi] Blanchard (p. 66)
[xvii] DoD RAM Guide (p. 3-10)
[xviii] Blanchard (p. 67)
[xix] DoD RAM Guide (p. 3-9)