Non-invasive Wireless Monitoring Provides Fast Payback

Figure 1. Steam Trap Monitoring:
Wireless data indicated serious
problems with a quarter of the traps.

The guiding principles we use to identify and execute continuous improvement projects include: • implementing performance-based maintenance;• reducing energy costs; and• increasing uptime. We identified several assets — steam traps and critical research freezers — for which a performance-based maintenance strategy could cut energy costs or boost uptime. Genentech has an annual steam-trap maintenance program. However, throughout the year, steam traps often fail. We estimate that steam loss from failed traps annually costs hundreds of thousands of dollars. So, monitoring to prevent or at least find failures as they occur could result in significant energy savings.

Figure 2. Crucial Freezer:
Amperage data provided by
monitor enabled detection of
imminent failure.

Research activities rely on ultra-low-temperature freezers. Reliability issues with older stand-alone freezers lead to downtime and potential loss of work product — mandating more-frequent maintenance to avoid these problems. Proactively detecting issues and preventing failures would result in increased freezer uptime and could potentially extend the useful life of the units. The ChallengesImplementing performance-based maintenance on existing systems raises issues such as cost and invasiveness. Asset-condition monitoring requires replacing analog instruments or installing new instruments to collect data. Traditionally, this work is invasive and demands a process interruption and engineering oversight. Furthermore, the new system must be commissioned for restart. This often includes inspections and leak checks as well as revalidation to return the system to service — frequently resulting in an unfavorable cost/benefit ratio. We identified a non-invasive monitoring solution from Cypress Envirosystems (www.CypressEnvirosystems.com), San Jose, Calif., that promised a cost-effective alternative. It uses devices clamped onto existing instrumentation to collect and wirelessly transmit data to a server for trending and alarming. This method eliminates the need for process interruption for installation, minimizes engineering oversight and offers other benefits (Table 1). However, it’s new technology and, thus, posed concerns about data integrity, wireless security and data integration into our existing infrastructure. To demonstrate proof of concept as well as to quantify potential savings, we partnered with Cypress Envirosystems to pilot the technology on some steam traps and freezers. Impressive ResultsWe began installing units in September 2008, putting them in service in our normal operating environment. Wireless Steam Trap Monitors (WSTM) were placed on 56 traps (Figure 1) at a cost of $42,000. The WSTM were intended to provide early indication of steam leakage or blockage, which leads to failure. Data were collected on a “blue box” server for trending analysis.

Figure 3. Low-Stage Compressor Amperage: Profile of abnormal freezer
indicates a potential refrigerant leak.

Fourteen steam traps showed signs of failure. The maintenance team investigated and replaced the traps. Early detection saved an estimated $53,000 in annual steam loss. Project payback was 10 months. Wireless Freezer Monitors (WFM) were installed on 20 critical freezers (Figure 2) $20,000. The WFM measure parameters such as low-stage and high-stage compressor amperage, internal chamber temperature and door open/close status, with the aim of giving an early warning of refrigerant leaks or door seal issues that lead to freezer failures. Four freezers showed signs of distress. Their temperature profiles indicated no problems with freezer functionality — but their compressor amperage profiles pointed to imminent freezer failure (Figures 3 and 4). The maintenance team investigated and resolved the issues without having to replace the freezers. Early detection saved an estimated $20,000 in replacement costs. Furthermore, we repaired the freezers before they failed, minimizing the risk of losing any research samples and, thus, protecting the product pipeline. Project payback was immediate. A Successful StartThe pilot projects demonstrated that the non-invasive wireless monitoring solution was a successful tool for implementing a performance-based maintenance strategy. Installation didn’t disrupt plant operations. No issues were found with wireless security. The data collected were easily integrated into our existing infrastructure. Overall results showed a less-than-one-year payback period — and certainly point to more extensive use of the technology for such applications and even more broadly.

Figure 4. High-Stage Compressor Amperage: Data for abnormal freezer
points to an electrical problem.

Since the initial pilot projects, Genentech’s effort now has expanded to include monitoring of HEPA filters on backup diesel generators using Cypress Envirosystems’ Wireless Gauge Readers. In addition, our site in Oregon has installed the Wireless Stream Trap Monitoring solution first deployed in our South San Francisco facility. Chris Stubbs is senior director, corporate facilities services for Genentech, Inc., South San Francisco, Calif. E-mail him at[email protected].