Faced with ever increasing competition on price for their products, chemical companies are constantly searching for ways to cut costs across their operations. This, in turn, puts persistent pressure on engineers to reduce both capital and operating expenditures — without compromising their plants’ reliability and performance in any way. Easier said than done, perhaps, but this is precisely what a new approach to the task of delivering process samples to analyzers actually delivers, claim its proponents.
Since coming into being some five years ago, the New Sampling/Sensor Initiative (NeSSI) has become the driving force behind the move to modularize and miniaturize process sampling systems. Now operating under the sponsorship and umbrella of the Center for Process Analytical Chemistry (CPAC) at the University of Washington, Seattle, NeSSI first surfaced as an ad hoc group of people drawn both from equipment manufacturers, keen to adopt the modular approach, and operating companies prepared to put the vendors’ prototype products to the test on their plants.
The initial concept was to build on a standard for modular surface-mounted systems developed for the semiconductor industry. This was modified for use in the oil, chemical and petrochemical industries by the SP76 working group of ISA, Research Triangle Park, N.C., and its resulting design became the basis for the ANSI/ISA-76.00.02 standard in 2002. This international standard lays down the specifications and dimensions for the sampling system substrate and the modular components — for sample flow control, conditioning and analysis — that plug into it. A modular sampling system along these lines consists of a series of close-coupled, 1½-in. square manifolds, each carrying one of the components fastened to the substrate with just four hex (Allen) screws.
Motivation for modular
Compared with traditional sampling systems, this approach seemed to offer significant savings. These were first quantified in 2000 by a team from ExxonMobil that looked at both the “cost of build” and “cost of ownership” of conventional process analyzer systems. Considering project data from around the company’s sites, the team concluded that 38% of the initial outlay was spent on the analyzer, 30% on the sampling system and 27% on providing a controlled environment or enclosure for these. However, if plants switched to the modular approach espoused by NeSSI, the team forecast that 40% savings could be achieved on the overall cost of build by reducing the expenses for the sample transport and conditioning system and eliminating the need for climate-controlled housings. Moreover, it projected a 35% saving in the cost of ownership through a combination of reduced sample volumes — and therefore smaller volumes of carrying and purge fluids — and easier maintenance and support, requiring fewer technicians (or enabling more analyzers to be supported per technician).
Jeff Gunnell, lead specialist for process analytics with ExxonMobil Chemical Company at Mossmorran, Fife, Scotland, was part of that seminal team that helped start spreading the NeSSI message. Asked where the initiative is now, he says: “After the field experiences with ‘Generation I’ NeSSI systems, we now have a solid platform to build on. There has been one round of iteration [in design] since 2001 and I would say the systems have been shown to work fine in the field within a realistic operating window.”
ExxonMobil has a total of 30 modular systems installed across several of its major locations (on two sites at Baytown, Texas, and at Sarnia, Ontario), sampling a variety of hydrocarbon process streams. The systems are currently approaching approval as “Best Practice” within ExxonMobil and that, says Gunnell, “means we are considering them for some major projects — and that could be huge numbers of systems.”
ExxonMobil’s modular sampling systems have come from two of the three main vendors involved in supplying the substrates and other flow components for ANSI/ISA-76.00.02-compliant systems. These are the Instrumentation Division of Parker Hannifin, Jacksonville, Ala., and Swagelok, Solon, Ohio, with Circor Instrumentation Technologies, Spartanburg, S.C. rounding out the trio.
All three companies have been making steady inroads into the process analytical market with their respective systems, reporting a doubling of sales each quarter at a recent CPAC/NeSSI workshop in Seattle. Commenting on the meeting, Swagelok’s manager for marketing resources, Dave Simko, says: “There is a consensus that NeSSI has arrived. Those companies using NeSSI systems have shown that the cost savings are real. Although component cost is higher, savings in design time and manufacture give a net saving — 30% lower in terms of overall cost.”
Swagelok’s MPC (Modular Platform Components) system (Figure 1) consists of a range of fluid-control components — such as shut-off, needle, metering, toggle and check valves, as well as filters — mounted on top of a substrate layer of 1½-in.-square modules containing specialized channel and flow components. Describing one of the first demonstration systems, Simko explains: “One system, which measures ppm H20 and O2 in a high-purity hydrocarbon stream, was built to demonstrate that the modular concept was feasible and practical. The complete system measures 24.6 × 75.2 ×19 cm (approximately 10 × 20 × 7.5 inches), which is considerably smaller than traditional panels and enclosures.”
Not only are the systems smaller but so too is the design effort required. Both Swagelok and Parker now have software packages that enable users to quickly and easily configure complete systems and generate bills of material and final assembly drawings.
The benefits reported at the Preem Refinery (formerly Scanraff) at Lysekil, Sweden, are typical. Here, Parker’s Intraflow system (Figure 2) has been implemented on a new flow-control system for dewpoint analysis. The field-mounted analyzer station monitors the moisture content of propylene gas in the refinery’s tanker-loading jetty area. Tony Carlsson, the refinery’s instrument engineer who specified the system, says: “This new surface-mount technology substantially reduces the space required to assemble analyzer stations, as well as the sampling volumes of the system. The plug-together nature of the Intraflow substrate greatly simplifies routine maintenance such as the replacement of filters. It’s a very attractive solution and I expect to adopt it on future installations and upgrades.”
According to Parker, a major advantage of its Intraflow substrate is intrinsic support for the three-way flow paths required for sampling systems. Unlike the linear substrates used for semiconductor “gas stick” delivery systems, Intraflow’s flow paths are all in a single plane and no manifolds are needed on additional substrate layers to implement the more sophisticated functionality of sampling systems.
In Swagelok’s MPC system, says Simko, sequential flow components define the flow path through the system. A drop-down flow component allows different sample streams, purge gases, flushing solutions and calibration or validation fluids to be introduced into the main flow path through the substrate.
With Circor’s μMS³ “micro Modular Substrate Sampling System” (Figure 3), on the other hand, the main flow path is actually external to the Lego-like substrate building blocks through what it calls NuBlu tubesets mounted on to the building blocks. The company’s process industry specialist, Robert Sherman, explains that these pre-cut and electron-beam-welded tubesets can be assembled, and reassembled, with the building blocks into virtually any configuration. Indeed, by keeping key substrate parts on-site, companies can rapidly modify any system. The external flow path also allows for visual flow validation and logical troubleshooting, he maintains.
The design procedure for the “build to order” μMS³ systems also is quite straightforward. A customer simply provides Circor with a flow schematic of the analyzer system. From this, Circor creates a manifold flow schematic and a component system drawing for customer approval. Alternatively, the customer — end user, integrator or analyzer company — can order the simple substrate components and quickly assemble its own systems.
The next generation
As ExxonMobil’s Gunnell says, the modular-sampling-system platforms are clearly in place and are acquitting themselves well at a host of sites around the world. These represent the Generation I NeSSI systems, but we now are already well into the Generation II era. This involves moving beyond the physical fluids-handling aspect of sampling to the ultimately more important issue of integrating the modular systems into the plant control hierarchy.
Another active member of the NeSSI steering group is Rob Dubois, senior analytical specialist for Dow Chemical at Fort Saskatchewan, Alberta. Systems from all the main vendors are in use at that site on ethylene-hydrocarbon and ethylene oxide production applications, as well as on a continuous-emissions-monitoring system for NOX. “Technically,” he says, “the NeSSI systems work well and many of the modular components which previously had been missing are now available or are coming to market. However, implementation has been slower than hoped for due to higher costs when compared to conventional systems. Right now, we are missing a ‘single block’ combined flow transmitter/indicator which would really simplify and ease construction.”
Commenting at the Spring CPAC/NeSSI workshop on the control aspects of sampling, Dubois said: “Analyzer systems might have become ‘smart,’ but the sampling system remains archaic. Today the industry uses a hodge-podge of analyzer I/O [input/output], PLCs [programmable logic controllers], DCS [distributed control systems], databases and proprietary systems.”
Generation II NeSSI systems will address this problem via a Sensor/Actuator Manager (SAM) operating system, which will act as the communication channel between any DCS and any analytical system. At the higher level, NeSSI systems communicate to the DCS using the OPC protocol over TCP/IP Ethernet, while at the field level Foundation Fieldbus (FF) has become the preferred route.
After two years of experimentation, the CPAC/NeSSI committee came to the consensus that FF was the right architecture for the project, including the need for intrinsically safe operation; it was duly incorporated into the Generation II draft specification in February this year. According to CPAC director, Mel Koch, “This initiative will help the automation industry adapt the merging class of ‘lab on a chip’ sensors to a miniature/modular ‘smart’ manifold.” Dubois is similarly enthusiastic. “Currently our systems are operated on a local level,” he says, “[but] once the intrinsically safe NeSSI bus is commonplace, and supporting sensors and actuators are available, you will see tighter integration of the sample systems.”
Not everyone agrees, however. Steve Broy, engineering director of Teledyne Analytical Instruments, City of Industry, Calif., contends: “The selection of Foundation Fieldbus in my view may not have been the best overall. This may limit the number of players going forward and might serve to limit proliferation due to the cost and complexity of implementing FF.”
That said, Teledyne is nevertheless one of many analyzer companies providing components for ANSI/ISA-76 compliant systems. Broy says, “We offer trace and percent oxygen sensors with signal conditioning and data acquisition on the substrate and thus far we have sold several sets in the past two years.”
Dubois adds, “Some major analytical vendors are now embracing the NeSSI vision and are adopting both the fluid handling components as well as the NeSSI-bus/SAM. Actually we have two possible buses: a miniaturized version of FF, which is work in progress, as well as an upcoming version of CAN [controller area network bus] being undertaken by the NIST IEEE 1451.6 effort.”
Other analyzer vendors involved include Panametrics, Waltham, Mass., and Bühler, Winnipeg, Manitoba, also with oxygen analyzers; X-ray Optical Systems, East Greenbush, N.Y., with an on-line sulfur analyzer (Figure 4); the Rosemount Instrument Division of Emerson Process Management, Austin, Texas, with pH and conductivity sensors; and mass flowmeters from the Fischer & Porter division of ABB, Norwalk, Conn. Honeywell Sensing and Control, Freeport, Ill., is part of a collaborative team, including Parker, looking at ways of developing new ranges of micro-scale analyzers and sensors that could be used in NeSSI systems.
There are limits, though. As Gunnell says, “You are not going to get the manufacturers of ‘battleship-sized’ GCs [gas chromatographs] to change but a modular sampling system can still sit side-by-side with the analyzer.” Teledyne’s Broy agrees: “Some measurement technologies will never be miniaturized in our lifetime. Right now it’s about putting sensors on a miniaturized platform and that forces you to examine how you have being doing things in the past, which is a positive exercise that has had some wonderful collateral fallout for us.”
Where does NeSSI go from here? Generation III systems undoubtedly in time should be able to offer additional functionality such as Koch’s “lab on a chip” and wireless communications, making them even easier to install and operate.