It’s not just progress in the technology that is spurring the acceptance, notes Roy Muston, senior automation consultant with Shell Global Solutions (U.S.) Inc. “The [drive] to run process units at maximum capacity and higher yields has increased the need for process analyzers,” he explains. At the migration’s heart lies technology transfer. The telecommunications indus-try and low-cost lasers for spectrometric use drive that transfer says Terry McMahon of PAI Partners, Leonia, N.J. He specifically points to the role that small, rugged telecom-munications spectroscopic components are finding as process analytics tools.
McMahon cites in particular Axsun Technologies, Billerica, Mass. That firm of-fers the IntegraSpec XL micro-spectrometer for general process applications and the In-tegraSpec XLP micro-spectrometer for meeting U.S. Food and Drug Administration’s process analytical technologies (PAT) requirements. It already is being used is in the dry-ing process for active pharmaceutical ingredients to meet PAT monitoring goals.
Encoded-photometrics infrared (IR) spectroscopy also offers promise. “Every-body is making a big deal about it, but it’s way too early to determine its impact on the chemical industry,” cautions Menlo Park, Calif.-based Steve Walton of PAI.
Even so, Aspectrics, Pleasanton, Calif., with its MIR-1 encoded-photometrics IR spectroscopy analyzer, grabs his attention. “Encoded IR mimics FTIR (Fourier Transform IR) spectroscopy,” Walton explains. FTIR typically has been used in the laboratory to measure organics through absorption of various infrared-light wavelengths. Encoded IR’s claim is that it attains the resolutions and separations normally found with FTIR but does so online, he adds. One potential use is for ambient air and stack-gas monitoring — for example, hydrocarbons speciation in homolog gases such as butane, pentane, etc.
Other technologies garnering greater interest include process mass spectrometry, including FT mass spectrometry (FT MS), which also is called FT-ion-cyclotron-resonance MS (FT-ICR MS), and nuclear-magnetic-resonance online processing, which also is known as industrial magnetic resonance. Meanwhile, a new platform for sampling is attracting attention. Dubbed New Sampling/Sensor Initiative (NeSSI), it combines modularity with micro-analytics and intelligent sensing and communications.
However, Walton notes the chemical industry is conservative, a relatively slow adopter. “They’re not going to put in anything that will undercut their operations — or something they can’t maintain.” Safety, reliability and cost are the biggest current chal-lenges for process analytics, adds McMahon.Nevertheless, these and other technologies are increasingly capturing industry’s fancy and promise to fundamentally change process analytics.
Minutes, not hours
The just-released FastGC (Figure 1) from ABB’s Process Automation Division, Lewis-burg, W.Va., certainly aims to play a significant role in continuous online analysis. “It does simulated distillation analysis of refinery crude and products,” explains Damian Huff, product manager. Laboratory analyzers now used to verify liquids’ distillation pro-files typically require two to three hours from sample collection to analytical results, he estimates. “Our instrument gives results every three to five minutes.” FastGC, which Huff says has a 99+% operating availability, fully networks with Ethernet, so real-time data can be displayed at operators’ workstations.
Figure 1. This unit provides results in 3 to 5 minutes, not the hours typically required.
The instrument can handle analyses for many refinery operations. However, it’s not suitable for heavy distillates, Huff notes. “We typically will measure liquids with a final boiling point of 470°C. It’s difficult to measure anything with a higher boiling point because it’s difficult to vaporize the sample,” he explains.
One potential use of an instrument like ABB’s is to monitor pipeline blending, in which refiners use the run-of-pipe from tank farms to transport vessels for final blending. This approach promises to save on tankage as well as offer better control of the blend but depends upon fast analyses. “Refiners produce fuels and put them into a blending tank. But because their old laboratory-based distillation analyses were [15-30 min.] aged, they could not realize the benefits of pipeline blending,” Huff explains.
Magnetic-resonance-imaging-based technology like that used in hospitals also is finding its way into plants. Invensys Process Systems, Foxboro, Mass., offers Process Magnetic Resonance Analysis (MRA) (Figure 2), which can analyze the composition of crude and final products at refineries, as well as streams at chemical and pharmaceutical plants, says Joe Fillion, who is responsible for worldwide sales and marketing.
ConocoPhillips installed a Process MRA at its 185,000-bbl/d refinery in Trainer, Pa., to analyze diesel and kerosene streams from two crude-oil distillation towers. “With some things we were measuring, we didn’t feel like we could [do those well] with in-ferred properties,” explains Tim Lowell, plantwide control systems engineer. “On our kerosene stream, naphthalene was measured with a smoke test,” he adds, noting he’s never spoken to anyone who’s ever successfully used that test. “That was the main thing that moved us to get the analyzer.”
“Before the Process MRA, we had essentially no real-time information,” Lowell stresses. Having the instrument has meant faster decision-making. “And we’re able to implement advanced control on the crude towers,” he adds.
Equistar Chemicals uses a Process MRA at its Corpus Christi, Texas, site to measure 57 components — normal parrafins, iso-parrafins, naphthenes, aromatics, ini-tial/intermediate/end distillation boiling points, etc. — every 12 to 15 minutes in a heavy feedstock stream at its cracking plant, to provide data for an optimization program. The unit dramatically reduced laboratory validation samples. “We take two lab samples a month. Before we took one to two daily,” says Tripp Howse, analyzer technician. Rival near-IR and mid-IR technologies could eat up eight hours from sample col-lection to results, Fillion says. Other advantages the Process MRA offers are distinct and linear spectra. Having linear spectra means orders-of-magnitude fewer samples to build the partial-least-squares models that confirm results, he explains.
Figure 2. One refinery is now using this technology to analyze 57 components.
Also, uptime can be an issue with near-IR and mid-IR, Fillion contends, with availability sometimes running less than 50%. In contrast, Howse says the Process MRA unit has had 99.5% uptime since it was installed in 2002.
“Daily, I check it for flow and temperature. Then I check the electronic side to make sure the indications the analyzer are giving are valid,” Howse explains. Weekly, for an hour, he takes the unit offline for sampling-system cleaning. “But essentially, the in-strument runs 24/7. It’s been over three years since I’ve had to do anything major.”
Uptime has been essentially 100% at ConocoPhillips’ Trainer facility. There, the Process MRA has achieved one continuous run of 396 days since it went online in 2003, says Lowell. “The only time we shut it down is to do routine filter cleaning. I don’t count it as being down,” he explains, noting the 12-hour-cleaning downtime occurs between scheduled laboratory samples.
Small but significant edge
Another promising technology is FT-ICR MS. It offers the possibility of separating spe-cies that are less than one mass number apart, McMahon says.
“An FT-ICR MS allows the end-user to avoid needing a GC to separate compo-nents in a gas-phase mixture,” notes Patrick R. Jones, a chemistry professor at the Uni-versity of the Pacific’s College of the Pacific, Stockton, Calif. The college purchased a high-resolution Quantra FT-ICR MS from Siemens Applied Automation, Bartlesville, Okla., for teaching and research.
“It has not been possible to teach graduate students, let alone undergraduates, about FT-ICR MS because of the high cost, delicate nature of the instrumentation and difficulty of use. The Quantra has addressed all those,” notes O. David Sparkman, ad-junct chemistry professor and manager of the college’s mass spectrometry facility. “The instrument has been running now for over two years with no maintenance,” Jones adds.
“The high resolving power of the instrument allows molecules of the same nomi-nal mass…to be readily distinguishable… without using chromatography,” Jones ex-plains. “That greatly simplifies the [analytical] situation.”
The technology (Figure 3), which Siemens claims can analyze off-gas samples in approximately 15 seconds, can replace two or three regular GCs or quadrupole MSs, as-serts product manager Wayne Rimkus.
Figure 3. Acrylonitrile plants already rely on this compact unit.
The transportable unit already is popular in acrylonitrile production, where sam-ples may have large amounts of water and ammonia, which can create difficulties for GCs. There also are several isobaric interferences — i.e., same nominal mass, but differ-ent exact mass — that a standard quadrupole MS cannot resolve.
“Catalytic fines can cause clogging and the hydrogen cyanide/ammonia mixture can polymerize,” Rimkus notes. The device can sample at pressures as low as 10 torr, which significantly contributes to its success, he believes. “The sub-ambient pressure pre-cludes much of the maintenance normally required, as well as avoiding most of the po-lymerization.”
At its Greenlake Complex in Port Lavaca, Texas, INEOS Nitriles has installed two Quantras. The company chose the units over GC, laser, IR-spectroscopy and other MS technologies, notes Michael Hoffman, facility analyzer specialist.
“The Quantra has performed as stated by Siemens,” Hoffman says. There were some initial issues with valve temperatures and sample introduction. “However, most analytical issues initially observed were related directly to learning to use this piece of equipment.”
Process analysis is benefiting from more than developments in analyzers. Sampling sys-tems are evolving, too, particularly thanks to NeSSI.
Making analyses simple and standardizing sample-system design is NeSSI’s goal, says Executive Director Dr. Melvin Koch of the Center for Process Analytical Chemistry (CPAC), based at the University of Washington, Seattle. “There’s also a huge opportu-nity to adapt the emerging class of ‘lab on a chip’ sensors to a miniature/modular ‘smart’ manifold which could fundamentally change the way industry does process analysis.”
“A smart system would employ a simple plug-and-play bus-connection sys-tem, eliminating the need for clunky conduit or cabling systems in electrically hazardous environments,” adds Dow Chemical’s Ron DuBois, who is based in Fort Saskatchewan, Alberta.
NeSSI, which is a non-affiliated, international, ad hoc group composed of indi-viduals, vendors and end-users, operates under the aegis of CPAC. Some participants in-clude Dow, ExxonMobil, ABB, Siemens and Emerson Process Management. “Swagelok, CIRCOR and Parker-Hannifin are the main platform providers,” Koch notes.
NeSSI’s modular platform comes via the SP76 Standard developed by ISA, Re-search Triangle Park, N.C., notes DuBois, explaining that ISA borrowed the platform from the semiconductor industry and then developed a standard for it. “This is one of the few cases where the migration is from the process environment to a laboratory/micro-reactor environment,” he adds.
What process analytics could the platform carry? The sky apparently is the limit, suggests Koch. “It [the process analytical device or system] has only to interface with one of the modules on the NeSSI platform.” He adds that several spectroscopic and separa-tion-based technologies have been demonstrated and commercialized in this configura-tion.
Koch emphasizes that NeSSI’s standardization promises benefits such as reduced installation and maintenance costs. With that platform and smart sampling devices, indus-try stands to gain better performance and reliability from analytical systems.
“[A key benefit is] the ability to use the fluidic platform and the electrical bus as tie-in rails for the emerging generation of micro-analytical devices,” adds DuBois. “Micro-analytics allows us to move our analytical systems out of expensive analyzer houses and put them right beside the process lines. We call this ‘by-line’ analysis.” This could bring considerable cost savings, he believes.
NeSSI is poised to play a significant role at plants. For instance, at the latest (May) NeSSI workshop, it was reported that a Singapore petrochemicals-and-refining facility is planning to install several hundred NeSSI systems employing “Generation II” smart modular components. Siemens has been awarded the front-end engineering con-tract involving 300-400 sampling systems for a Singapore petrochemicals-and-refining facility. While the company is testing both NeSSI and traditional systems, it hopes that the project can provide the first broad deployment of NeSSI systems using Genration II controls. (For more information on NeSSI, see CP, Sept. ’05, p. 42.)
Fulfilling a need
Operating companies such as Dow, DuPont and ExxonMobil now drive the acceptance of NeSSI, Process MRA or other analytic technologies, say McMahon and Walton. “The companies have done enough to determine that there are significant economic gains by using the systems,” notes McMahon. So, it doesn’t take much analysis to predict that online process analytics will gain more favor and use.