Numerous previous limitations have been dealt with and at the same time prices have come down to, or in some cases even less than those of conventional level instruments. This article compares conventional technologies with the new radars and considers arguments to replace them with radar.
Radar in the past
Radar has been used in level measurement for several years and in a variety of applications. However, in the petrochemical industry it has been limited mainly to storage measurements, process measurements being performed with more conventional technologies. There were good reasons for this. Process conditions are, to say the least, rather 'instrument unfriendly' and previous generations of radar level instruments proved not to be sturdy enough to survive in those circumstances. The author had personally seen few radars in process measurements that lasted longer than six months without requiring major maintenance.
Since most radars were mounted without an isolation valve this meant stopping the process and quite frequently emptying and cleaning the vessel before the instrument could be removed. The best reason to apply radar (at that time) was that other instruments failed even sooner - or could not reach the required accuracy. The new generation of radar level instrument has changed that completely.
Improved process seal
Flange temperatures of -100 to +400°C, at pressures from full vacuum to 10 MPa are now made possible by using waveguides made out of ceramic instead of PTFE. Not only does it allow for very high temperatures but also makes it possible to use hard sealing materials like tantalum or graphite, these being able to withstand high pressures and temperatures. These combinations can cope with almost 98% of all chemicals and mixtures.
This sealing material is not there to contain the process environment, that is job adequately done by the instrument's mechanical housing. It is there to protect the radiating antenna (the most vulnerable part of the instrument) against process influence. The major cause of radar breakdown has always been process fluid that passed the soft sealing and either damaged or short-circuited the antenna.
Improved electronics
The use of new electronic components dramatically lowered energy consumption, in so much that a two-wire, 4 to 20 mA connection is all that is needed. That includes HART communication, both in flameproof and intrinsic safe designs.
Simplified operation
While radar is still a very complicated technology, the new instrument itself is not. An on-board microprocessor does a lot of the 'tweaking' that previously required a skilled radar expert. Just connecting a PC onto the instrument wiring, setting the range, selecting the process conditions is all that is needed to have the instrument operational. The same function may also be performed with a standard HART handheld communicator, but that will take a little longer than using a PC.
Conventional measurement performance
Does it make sense to replace your existing instruments? This is a tough question for most instrument people and their budget guardians. Why should anyone in his right mind throw away something that appears to be working well?
Let us take a look at the most frequently used technologies, how they work and perform and what radar has to offer instead.
Differential pressure
DP (differential pressure measurement) is by far the most applied technology for liquid level measurement. It is low cost, easy to apply, accurate and it always works. At least that is its image, but does it live op to it in real life? Let us look at what it takes to create a safe, accurate and reliable DP level measurement.
* Fluid density
The first thing that must be appreciated is that DP does not measure level, it measures the force exerted by a fluid column. As long as the density of this fluid is known it is fairly easy to correct and recalculate this to real level. This is true as long as the density does not change. But liquid densities in a process vessel do change. Either the product changes or the temperature does, most likely resulting in a change in fluid density. And how is a DP transmitter supposed to 'know' about this change? The truth is; it doesn't! The best solution is to add some form of density measurement and have the process computer or a multivariable transmitter correct for density changes.
* Gas density
If the process runs under elevated pressures then the transmitter not only has to deal with the 'weight' of the fluid column but also with that of the gas column above the fluid. Especially at higher pressures this can introduce considerable errors. The combined error can easily be 10% of span with 'ideal' gases but when super-compressible gases are involved things get worse. Densities become much higher and so does the error. An additional problem is that errors caused by pressurising the gas layer only show up in the instruments zero reading. So if the vessel is empty, but at working pressure, the transmitter is already indicating 5 or 6%. Since most operators don't like to start with an error reading, the obvious solution is to zero the DP transmitter. Later, when the level is raised, the gas column is replaced with liquid. The resulting force exerted by the fluid column is that of the 'weight' of the fluid minus the weight of the replaced gas. If the DP transmitter is not corrected for the gas density it will indicate a reading that will be too low.
* Safety concerns
The level in a vessel could be higher than is assumed. The operator may decide there is still 8% or more headspace left in the vessel, but in real life the upper tapping can already be flooded and the DP transmitter will not indicate further level increases. Then there are the errors resulting from the tubing connecting the transmitter to the vessel, the possibility of blocked tubing or valves. If there is a concern for safety, software should be included in the process controller to detect plugging, as most transmitters will not report it.
These error sources just discussed apply to all types of DP measurement, including remote sealed DP. Correction is still necessary for the effect ambient conditions have on the capillary fill fluid.
* Is it still safe to use DP?
Yes, all these shortcomings can be allowed for by installing additional instruments. For example: US patent 5.791.187 issued 11 August, 1998 (Chang, Hak Soo) describes in detail how to effect the necessary DP installation for measuring water level with a bubbler system. It involves three pressure and one temperature transmitter, a control scheme plus a calculating device. (Hydrostatic tank gauging uses a similar technique.)
Remember that these schemes either need a separate calculating device or each transmitter has to be hooked up to the control system - and that does not come for free. Another option is to live with the inaccuracies and range the transmitter so that the vessel will never overflow, even at a worst-case fluid/gas density combination. The price paid is that the vessel can, even in emergencies, only be used up to 70 to 80% of its capacity and the real level will never be known.
* Displacers
Another popular way of measuring level is using a displacer, especially for small ranges. It is not a real low cost measurement – but wide pressure and temperature ranges often make it an attractive solution. Just like DP it is not a 'real' level measurement. This technology depends on the Archimedian principle in that it senses the upward force exerted by a replaced volume. That makes it just as dependent on the fluid and gas properties as DP measurement. The displacer body is in permanent contact with the medium, which makes it unsuitable for any fluid that has sticking, abrasive and build-up properties. Boiling or dirty fluids present additional problems. Some of them can be dealt with by putting the displacer in a bypass tube, but these have their own set of problems. This is not the place to discuss the pro's and con's of bypass tubes but generally it is a bad idea to use them in processes where recipes are changed 'on the fly' or where reactive chemical rules apply.
* Floats
In contrast to the two previous technologies floats are 'real' level instruments. They are often combined with a magnetostrictive readout, which makes it an accurate measurement. If mounted directly in the vessel they are real low cost instruments, but then the temperature range is very limited and isolation is impossible. An alternative is to mount them in an insulated bypass tube and mount the magnetostrictive readout outside of the insulation. The float has to be designed carefully to fit the full density range of the fluid. Just like a displacer it is not suitable for sticking, abrasive, build-up and boiling fluids. The float and the guide are in permanent contact with the fluid and floats must be handled with care. If a float fails for whatever reason it always sinks, indicating low - without any warning.
* Long term cost of ownership (LTCO)
Purchase cost alone should never be the deciding factor in the selection of process control equipment. There is also cost of capital, planning, engineering, installation, commissioning and of course maintenance and spare parts. Some examples of how different technologies effect LTCO:
• Design it right the first time
Proper engineering for the technologies discussed above requires good knowledge of the fluid properties at the actual process conditions. Those are hard to come by at the design stage – it takes time to gather them and engineering hours are costly. After that, floats or displacers have to be replaced or DP ranges be recalculated during plant start-up because some things have changed after writing the order specification. In the case of radar, all that is needed are the general process properties to determine whether or not radar is suitable. Then, no matter what the actual process conditions will be, as long as they are within wide instrument limits, it will work.
* Less maintenance calls
Over the lifetime of an instrument, nuisance calls generally cost more than actual repairing or exchanging. Each time a technician has to check the proper operation of a transmitter the LTCO goes up. And over 40% of the time he will find nothing wrong. It must be kept in mind that whatever the outcome, the average cost for an instrument call remains - R500 or more.
Once accepted by operations, radar has proven to be a trustworthy instrument and the number of nuisance calls reduced sharply.
* Fewer spare parts
Spare part stocks should contain enough variation in range, size and materials to cope with all applications. Granted, one normally won't need that many displacer bodies in stock, they don't break as easy as floats or DP transmitter seals. But when stocked, floats and displacers have to be exactly sized for each application and for transmitters you need to have at least several range models and capillary lengths.
Radar comes in one range, no matter what the measurement range is; the instrument range is 32 m. Standardising the flange size would require just a small stock to cover all applications.
Conclusion
In many cases, radar is the best option. If the process fluid has a dielectric constant greater than 1,5 and there is a 4" (102 mm) flange on the top of the vessel, with a free 'line of sight' between that flange and the fluid surface, then radar is a good proposition.
Free line of sight does not mean that it must be free of obstructions, because radar can cope with many obstructions. It does mean that there must be a straight path available for the microwaves between the point where they leave the horn antenna and the fluid surface. In vessels with many microwave obstructions (stirrers, etc) the radar can be pointed down a tube or stilling well, provided there is a free exchange of fluid between the tube and the vessel. If the instrument must have a separating valve, a full-bore ball valve is mounted on top the tubing flange and the radar is installed on top the valve.
There is no concern over fluid density and changes, as radar always senses the fluid surface. The composition or behaviour of the gas phase is of virtually no concern. While one cannot flood upper tappings or displacer bodies, radar will go on measuring the distance to the surface, even when the fluid rises into the horn. It is not advised to run it that way, but it still works reliably.
If there is no room for a 4" flange to accommodate a horn antenna, one of the new rod antennas may be used. Although they are made out of PTFE they have an excellent process seal and the same measurement qualifications as the horn antennas. However, the use of PTFE does limit the pressure and temperature specs.
In view of the ever-increasing demands on safety, quality and cost competitiveness it really does make sense to review new and present level measuring designs. Radar is not a 'cure for all diseases' but, if the effort is made to understand it and then apply it well, it can make life a lot easier.
| Tel: | +27 11 795 3249 |
| Email: | [email protected] |
| www: | www.vega.com/en/home_za |
| Articles: | More information and articles about VEGA Controls SA |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version