Idc Engineers

Designing a Control System for High Availability

Art Pietrzyk, TUV FSExp, Rockwell Automation
Brian Root, Redundancy Marketing Manager, Process Initiative, Rockwell Automation
Paul Gruhn, P.E., CFSE, Training Manager, ICS Triplex


Introduction

When hearing the term “high availability,” many engineers think of redundancy as the only method for achieving higher availability. However,redundancy increases the number of components, which increases the number of potential component failures. Therefore, redundancy, if not applied properly, can actually decrease system availability. So, should redundancy remain top-of-mind or should alternate methods be considered?

This paper will discuss redundant and non-redundant methods for achieving high availability of control systems, aswell as improvements in control technology and recommended control system designs. The paper will also highlight features within the Rockwell Automation Integrated Architecture™ platform and ICS Triplex product lines that can help achieve higher availability.

Please note that this paper focuses on high availability for control systems. This paper does not address safety system concepts ordesign philosophies.

What is High Availability?

At the most basic level, availability can be defined as the probability that a system is operating successfully when needed. Availability is often expressed as a percent.
Expressed mathematically, availability is one minus the unavailability.

Availability (A) is calculated using the formula A = MTBF / (MTBF + MDT), where MTBF is Mean TimeBetween Failure and MDT is Mean Down Time. MDT is often assumed to be the same as MTTR, the Mean Time to Repair. MTTF, Mean Time To Failure, is often considered interchangeable with MTBF, although there are subtle differences. Another common term in the field of reliability engineering is failure rate (λ) which is expressed as 1/MTBF.

The term high availability has been used to encompass all thingsrelated to productivity, including reliability and maintainability. So let’s take a closer look at these terms as well.
Reliability

Reliability can be defined as the likelihood that a device will perform its intended function during a specific period of time (often called the mission time). It is a measure of system success over a time interval.

To help make sure that products meetcustomer expectations, reliability can be designed using techniques such as Component Derating and Design through Six Sigma.
Diagnostics are necessary in order to detect faults and alert personnel when faulty hardware needs to be replaced. This helps achieve high availability. The level of diagnostics for the Allen-Bradley® ControlLogix® programmable automation controller (PAC) from RockwellAutomation exceeds 90 percent, meaning most failures can be detected and appropriate actions taken. ICS Triplex Trusted™ and AADvance systems have diagnostic coverage in the range of 99 percent.

But even the most robust and reliable system may not be the most available. To be available, a system must also be easy to troubleshoot, modify and repair during the mission time, which may exceed a decadeor more.

The Impact of Maintainability on Availability

Maintainability is the ability of a system to be changed or repaired.

Factors that affect maintainability include:
• System and component-level diagnostics for detecting and isolating failures
• Annunciation of faults
• Tools for troubleshooting
• Trained personnel
• Accessibility
• Time to replace or repair
• Ability to addcomponents or make changes

Maintainability of a system significantly impacts the end user’s perception of availability. For example, today’s automobiles have diagnostic features that can improve availability (e.g. tires that heal and/or monitor pressure, electronic ignitions, diagnostic message displays that indicate a malfunctioning system and the ability to call for roadside assistance)....