TM 5-698-1 Reliability/Availability of Electrical & Mechanical Systems for Command, Control, Communications, Computer, Intelligence, Surveillance and Reconnaissance (C4ISR) Facilities (19 January 2007)

Overview

TM 5-698-1 gives facility managers the information and procedures needed to baseline the reliability and availability of their facilities, identify weak links and choose cost-effective ways to improve them. Its scope is the electrical and mechanical systems of C4ISR (command, control, communications, computer, intelligence, surveillance and reconnaissance) facilities, which in practice resemble military data centres. It supersedes the edition of 14 March 2003 and reflects the move toward commercial practices.

Chapter 1 defines the basic terms: availability is the proportion of time a system is ready to perform its required function, written as 0.99999 or 99.999 percent; reliability is the probability of no failure; maintainability is the ease of restoring an item. The text notes that the electric power industry has historically used reliability to mean availability. Chapter 2 covers probability, reliability calculation and availability calculation (block diagrams, series and parallel structures, MTBF and MTTR combinations).

Chapter 3 describes the improvement process on two tracks. For new facilities: set the availability requirement, derive reliability and maintainability requirements, develop one-line diagrams, predict by analysis and test. For existing facilities: collect data, measure baseline performance, identify shortfalls and rank improvements by cost-benefit analysis. Redundancy and reliability-centered maintenance (RCM) are treated as improvement tools. Chapter 4 presents the modelling methods (minimal cut sets, network reduction, Boolean algebra, Markov, Monte Carlo) and worked examples.

The central warning is that results are extremely sensitive to assumptions, method and data, and that no result should be distributed without documentation of those attributes. The text also notes that UPS battery life was historically assumed to be limitless, that standby equipment has different failure distributions from primary equipment, and that upstream effects of failures, such as a breaker that fails to open, are missed by simple top-down models. The manual dates from 2007; the methods remain sound, but component data should come from TM 5-698-5.

Key points

1. Availability is the fraction of time a system can perform its function, written for example as 0.99999 or 99.999 percent (five nines).
2. With exponentially distributed failures, reliability at an operating time equal to the MTBF is only about 37 percent (e^-1 = 0.368).
3. Inherent availability Ai = MTBF / (MTBF + MTTR) x 100; a 95 percent requirement might be met with an MTBF of at least 600 h and an MTTR of at most 3.5 h.
4. Observed availability A = uptime / (uptime + downtime); 10 h of downtime gives 95.83 percent over 240 h but 99.86 percent over 7,200 h, so contracts must state the measurement interval.
5. A five-subsystem example (MTBFs of 750-4,000 h, MTTRs of 1-4 h, some duplicated) yields 99.7236 percent availability over 10,000 Monte Carlo trials of 1,000 h; the lowest-MTBF subsystem contributes 32 percent of system unreliability even with dual redundancy.
6. An existing facility has three ways to raise availability: selective redundancy that removes single points of failure (generators, chillers, fuel supply), RCM to optimise maintenance, and replacing low-reliability subsystems.
7. For a new facility the sequence is availability requirement, derived reliability and maintainability requirements, one-line diagram, predictive analysis, testing, design revision and continuing assessment in operation.
8. The reliability of a series system is the product of component reliabilities (Lusser's law); the eight-step analysis process in chapter 4 requires validation throughout.
9. For simple systems cut-set, network reduction and Boolean methods give similar results when assumptions are equal; larger systems and time-dependent failures (belts, valves, batteries) call for simulation.
10. When one redundant unit fails it must be repaired or replaced to keep the intended redundancy, which is the basic argument for a maintenance programme.

Using it in design

• When writing an availability target into a data centre specification, state the measurement interval as well; the same downtime gives 95.83 or 99.86 percent depending on the interval.
• Convert the user-facing target (operational availability) into controllable design measures (MTBF, MTTR, inherent availability) and allocate targets to each subsystem.
• Model power and cooling topologies with block diagrams or Monte Carlo; find which subsystem contributes most to unreliability (32 percent in the example) and concentrate redundancy there.
• Do not assume unlimited UPS battery life, and treat the different failure distribution of standby equipment and failures with upstream effect, such as a breaker failing to open, separately.
• Document the source and uncertainty of the MTBF and MTTR data used as model input; the text names the data as the single point of vulnerability of the model.
• In maintenance planning remember that redundancy reduces the probability of system-level loss but does not reduce the total number of failures.

Who should read it

Design and operations engineers who set availability targets, compare topologies and carry out RAM modelling.

Document contents

1. Chapter 1 Introduction: purpose, scope, definitions, historical perspective
2. 1-6 Relationship among reliability, maintainability, and availability
3. 1-7 The importance of availability and reliability to C4ISR facilities
4. 1-8 Improving availability of C4ISR facilities
5. Chapter 2 Basic Reliability and Availability Concepts
6. 2-1 Probability and statistics
7. 2-2 Calculating reliability
8. 2-3 Calculating availability
9. 2-4 Predictions and assessments
10. Chapter 3 Improving Availability of C4ISR Facilities
11. 3-2 New facilities / 3-3 Existing facilities
12. 3-4 Improving availability through addition of redundancy
13. 3-5 / 3-6 Improving availability through reliability-centered maintenance (RCM)
14. Chapter 4 Assessing Reliability and Availability of C4ISR Facilities: methods, modeling examples, modeling complexities
15. Appendix B The Mathematics of Reliability
16. Appendix C Point Estimates and Confidence Bounds
17. Appendix D Factors Influencing Field Measures of Reliability

Read the document

Publication details

Authors

Headquarters, Department of the Army

Publisher

Headquarters, Department of the Army (U.S. Army Corps of Engineers)

Year

2007

Document language

English

Source type

Government / public

Access

Open licence — downloadable

Licence

U.S. public domain — U.S. Government work; free to reuse

Licence evidence

www.copyright.gov

Citation