COLOR MEASUREMENT SOLUTIONS

A drifting reading in an inline optical colorimeter is almost always due to two factors:

Coating/fouling forming on the optical window surfaces. Fouling cuts down the amount of excitation light entering the fluid, therefore reducing the amount of light being received by the instruments’ photodetector. Because colorimeters use light absorption as the basis for their measurement, a falling light level indicates a deeper color score. As fouling continues to accumulate, the color score continues to increase – the drift you are seeing. Using a referenced based measurement (i.e. dual wavelength vs. single wavelength) can resolve this issue. In a dual wavelength measurement, an NIR reference wavelength is used to monitor the coating on the windows and the amount of absorbance it causes. As fouling creates absorbance for both the measurement wavelength and the NIR wavelength similarly, subtracting ABS(NIR) from ABS(COLOR) will correct the color absorbance signal to a give true color reading. Note that it can be important to establish any absorbance ratio between ABS(COLOR) and ABS(NIR) and weight the subtraction accordingly. For instance, if the window fouling causes half the absorbance on the NIR reference channel as it does on the COLOR measurement channel, to get true color correction, it is important to weight the subtraction so that ABS(TRUE COLOR) = ABS(COLOR) – (2 x ABS(NIR)).

Lamp aging - In older optical colorimeters, an incandescent bulb is often used to generate light for the measurement. However, incandescent bulbs are constructed with metal elements and over time, these elements evaporate due to their temperature and coat the inside of the bulb, making their output dimmer and yellower. Modern systems use LEDs as their light source and these resolve the aging lamp issue because the output can be automatically held constant and does not “yellow” over time.

Optical colorimeters pretty much all work using absorbance as their physical measurement. To measure a yellow color, you need to use a blue light source (yellow’s complementary color) and have a measurement path sufficient for the max. depth of yellowing you see. For detecting very light-yellow coloration in your process, you need a longer pathlength than if you need to measure down to deep yellow/brown coloration. In your case (and assuming there are no issues with the equipment itself), it is possible you need to adjust either the wavelength of light you are using for measurement and/or increase the optical path across which you are measuring. Changing wavelength is easily done in a Kemtrak unit as it takes only a few minutes to install a new LED. Optical path adjustment may mean modifying or changing your measurement cell.

Difficult flow stream conditions like this can mean that an optical measurement is not a practical solution. However, there are a few ways to tackle this problem:

1. Install a measurement cell with a water (or other liquid) jet system that will blast the coating off the cell windows when actuated. You should do this often so any coating does not have time to really establish itself on the window surface. The Kemtrak DCP007 colorimeter has a built-in automatic routine to do this and cells can be supplied with water jets.
2. Install a probe in a retractable probe housing (such as an Exner EXtract device) instead of a flowthrough cell. This will allow the probe to be withdrawn from the line, removed, cleaned and returned without having to shut the line down. In some cases, retraction and cleaning could also be automated, with the probe being withdrawn from the line every hour or so and automatically cleaned
3. Install the optical measurement sensor (inline cell or probe) in a bypass loop to allow isolation from the process. Once isolated, the measurement cell or probe can be removed from line and cleaned. After cleaning, reinstall and return to service.

Keeping an optical measurement online and reliable is achievable if pipe arrangements like a bypass are installed or automatic cleaning routines using equipment such as cleaning jets or retractors are implemented. Even where conditions are extreme, automated solutions can greatly improve the quality and consistency of color measurement while reducing the amount of maintenance required.

Measuring two different colors is possible using a dual wavelength colorimeter. In this case, one wavelength can be configured to measure the blue color and the other set up to measure red. The instrument will give two outputs – blue and red. During normal operation, the blue output will be trending and the red output will be at or almost zero. Should the chemical addition fail, then the trends will reverse and it will be possible to see the color change in real time. Alarm limits can be set around these trends to alert operators to the problem. Furthermore, the analyzer can be used to automatically switch the stream to your holding tank when red reaches a pre-set level, just in case the operators are handling something else at the time.

There are a few reasons why your colorimeter is non-linear:

1. One of the challenges of colorimeter designs is excluding stray light. As color increases, absorbance of the wavelength of interest increases and the ratio of this to stray (unabsorbed light) changes. At high absorbance values, this can make the measurement non-linear. Many inline colorimeters use a white light source (often an incandescent bulb) for the measurement and install bandpass filters to exclude all wavelengths except for the wavelength of interest. Bandpass filters are not perfect and their out-of-band blocking capability varies dramatically depending upon the type used. In addition, the “better” the filter out-of-band blocking, the lower its light transmission capabilities, and this impacts instrument range and performance. LED light sources are a better solution as the light they generate is only associated with the wavelength of interest, removing the need for wide out of band blocking as unwanted light is no longer there and greatly improving linearity across wide color ranges.
2. If your colorimeter is installed in PVC or fiberglass pipe, it is possible that stray light is getting into the measurement circuit through the pipe walls. Wrap the pipe 10 diameters upstream and 10 diameters downstream with a black covering to exclude this light coming through the pipe.
3. The optical pathlength is too long in your instrument. Optical pathlength must be set to ensure that the maximum absorbance at the maximum color you expect does not exceed the instruments measurement capability. However, this can be a compromise as shorter pathlengths will have a lower resolution when it comes to very low color levels.

If stray light is not a problem and the optical pathlength cannot be changed, a potential solution is to calibrate your instrument using a polynomial expression or a piecewise calibration table instead. Setting up these methods will help correct the non-linearity you are seeing and bring the online measurement in line with the offline sample results you see. Kemtrak colorimeters offer linear, polynomial, and piecewise calibration as standard.

An inline colorimeter does not need to be large. Many traditional colorimeter designs have electrical/electronic components installed within light source and detector housings, so installing in a hazardous area requires explosionproof (or at least intrinsically safe) versions. The smart solution is to use colorimeters that use fiber optic cables. Using fiber optics removes the need for electrical/electronic components at the sensor, so the sensor is a small cell that can be fitted into your existing line (just like a pipe fitting) and connected to dedicated colorimeter electronics mounted remotely. Fiber optic cable runs can be very long (up to 100m), so explosion-proofing can be avoided if the colorimeter can be fitted inside a switch room or control area where ordinary area equipment is permitted.

Aeration is the number one interfering factor in colorimeter performance. In optical instruments like colorimeters, entrained air/gas will make the measurement noisy and create errors. Extreme aeration can cause measurement failure. Sometimes the issue is other equipment in the line and sometimes it is the process itself. There are several strategies that can be used to improve measurements being disturbed by aeration, things like:

• Suppressing aeration in the flow stream by increasing line pressure
• Remounting the colorimeter inline measurement cell away from cavitation generators such as valves and other pipeline components
• Repairing pump seals.

If all else fails, a sample and hold strategy in a side stream can be utilized. Flow is allowed to pass through the measurement cell to ensure it is representative of the main line. The sample stream is then stopped and held in the colorimeter measurement cell for a short period to settle before measurement. Flow is then started again and the cycle repeated.