In a perfect world, all the instruments used for process control would maintain their specified accuracy under any and all conditions short of damage, and maintain the same accuracy for the same process variable value regardless of what happens between measurements.
In the real world, instruments’ characteristics change, even if only slightly, as their operating conditions change. In sanitary processes subjected to clean-in-place (CIP) and sterilization-in-place or steam-in-place (SIP), instruments must withstand rapid changes in temperature and pressure while subjected to hot water, steam or aggressive chemicals. The temperature cycling involved can cause instruments like pressure transmitters to drift out of calibration, leading to out-of-spec product, missed time to market and possible recalls.
What is batch repeatability?
Quality control in batch production of products like pharmaceuticals and food can be a real challenge. Ideally, once a batch process is set up properly, every subsequent batch should be identical: the “golden batch.” This is not easy; the process spends considerable time starting and stopping, yet the end result must comply with strict government quality regulations. For such a system to succeed everything involved be repeatable within very close tolerances, and a small error in measurement can keep this from happening.
Consequences of failure
Several problems can arise from improper pressure measurements: tanks can overflow, leading to costly loss of product; chemical reactions can occur at sub-optimal conditions, leading to poor batch quality or batch loss; and inadequate filtration can occur, leading to wasted product or the need for additional processing. These problems impact production by reducing product quality and yield.
Failure to make each batch exactly the same can be a particular problem in the pharmaceutical industry, in which something as simple as an out-of-calibration pressure transmitter can cause the rejection of an entire batch. This means more than just a lost batch of product; it can entail loss of capital, delayed time to market, and the need to bring in more people to diagnose and correct the problem. And if the defective batch is actually shipped there is the possibility of a product recall with a cost that may reach billions of dollars—not to mention the potential damage to reputation.
The costs associated with failure can be described by the 1-10-100 rule. Essentially, the rule states that for every $1 a manufacturer spends to prevent failure, they would expect to pay $10 to correct a process during manufacturing and $100 to remediate failures after they occur. For the pharmaceutical industry, the costs of failure can be much higher due to government fines and litigation if a finished product must be recalled.
The challenges of CIP/SIP
The CIP/SIP processes used in hygienic applications can be a particular problem, because they entail extreme fluctuations in media temperature. Repeated large changes in process medium temperature can harm any sensing device; gaskets and other polymeric components can degrade, and different coefficients of expansion in the various materials that make up the device can cause not only calibration drift but even eventual failure. The frequency of such exposure can range from a few times per year to several times per day. A typical CIP process will include multiple steps with temperatures of up to 30°C (80°F) and at least one with temperatures up to 70°C (144°F) or 90°C (176°F) (depending on circumstances). Temperatures during SIP can be even greater: In a bioreactor, for example, SIP is often required to raise the coldest point in the system to 121°C (225°F) for 15 minutes. This will generally involve a gauge pressure of 1.2 bar (120 kPa or 17.4 psi). Some SIP processes can expose the transmitter to temperatures of 140 ºC (256 ºF) for shorter periods of time, and experience has shown that short-duration exposures to this sort of temperature have more effect on transmitter stability than longer exposure to slightly lower temperatures. Sooner or later the calibration will be so far off that the process will go out of spec.
For a more detailed look at what can happen, consider that a generic pressure instrument accurately zeroed should measure its first batch accurately. When the CIP or SIP process starts, the transmitter will inherently drift due to the quick process temperature rise, as shown in Fig. 1. When this process ends, the transmitter will try to return to 0 with minimal shift, but the drift from CIP/SIP will affect the quality of the sterilization process; the time back to zero can delay the process and affect batch quality; and the ability to return back to 0% of span affects the repeatability and quality of batches, as well as effects the time between calibrations. And the need to prevent these problems can mean more frequent calibration than normal.
Choosing pressure transmitters for batch-to-batch stability
The answer, clearly, is to be sure to use only pressure transmitters rated for this service, but that may not be as simple as it sounds. Many pressure transmitters are characterized for long term drift (often given in terms of percent of full scale range per year), but it may take some searching to find out under what conditions that long term drift specification applies. Long term drift testing done at a constant or near-constant temperature can mean little for a unit that will be subjected severe temperature cycling. It’s important when specifying a transmitter to be subjected to CIP/SIP to find out which models are designed and characterized specifically for such service.
When choosing a transmitter for a particular application, the first step is to be intimately familiar with that application. What pressures are involved? What temperatures, and for what duration? What chemicals? Once all that is known, the next step is to look for a transmitter that matches those particular conditions. In general, look for a transmitter that has been tested to 140°C, that has a diaphragm compatible with the chemicals to which it will be exposed (typically 316L stainless steel), and that has the appropriate surface finish for the application — typically 32 Ra standard per 3-A requirements, or 15 Ra per ASME-BPE requirements. It should have a crevice-free SST design for maximum cleanability.
CIP/SIP service can make it difficult to ensure batch-to-batch repeatability. It is important to choose instruments designed to maintain their accuracy despite the stresses that will be applied to them, to ensure a high quality result, and avoid the high cost of failure.