Wesley G. Poynter
Kam Controls Inc.
Houston
Roy J. Barrios
Louisiana Offshore Oil Port Inc.
New Orleans
The well known Karl Fischer method for analyzing water content of nonaqueous samples has been modified to facilitate field use. The conventional, visual endpoint titration has been altered to produce a coulometric, or electrochemical, procedure.
The coulometric method has all the advantages of the original Karl Fischer laboratory method, but can be packaged in a small, convenient module. The analysis takes little time and can be used in the field as well as the laboratory.
The relatively new coulometric method is described in API Manual of Petroleum Measurement Standards (MPMS) Chapter 10.9 and in ASTM D 4928. The titration is gaining increasing acceptance in the petroleum industry, where it is used by production engineers, transporters, and processors.
WATER ANALYSIS
For decades, the centrifuge method has been the accepted technique for measuring water content in crude oils and other hydrocarbon stocks. This method takes little time and is easy to perform. Sometimes, however, it is limited in its ability to measure water content accurately, particularly if the hydrocarbon stock has a high solubility for water.
When maximum accuracy is required, or when water solubility is high, a laboratory method called "water by distillation" often is used. This method is described in API MPMS Chapter 10.2, and in ASTM D 4006. The water by distillation method measures both dissolved and suspended water.
The water by distillation method is very accurate, but has two drawbacks:
- It is a laboratory method and is not suitable for use in the field, where timely evaluations often are required.
- It typically requires at least 2 hr, and sometimes 4 hr or more, for each determination.
In 1935, Karl Fischer, a German chemist, developed a titrimetric method for water determination. The method involves a chemical reaction between water and "Karl Fischer Reagent," which is a mixture of sulfur dioxide, iodine, pyridine, and methanol. The reagent reacts with dissolved and suspended water.
For many years, this method was limited to laboratory use because of the equipment and chemicals required to carry out the determination. But the newer, coulometric Karl Fischer method is faster and more convenient than these conventional methods for measuring water in hydrocarbons.
COULOMETRIC METHOD
Parallel tests have shown that the coulometric Karl Fischer method for measuring water in crude oils is at least as accurate as the API water-by distillation method. The time and portability advantages of the new Karl Fischer method offer convenience over the laboratory based distillation method, and its reliability exceeds the centrifuge method.
Louisiana Offshore Oil Port Inc. (LOOP) collected and published several sets of data comparing the Karl Fischer coulometric method with the water by distillation method. The data were published in the International] School of Hydrocarbon Measurement proceedings between 1988 and 1991.
Over a 5-year period, LOOP ran parallel tests using the coulometric Karl Fischer, distillation, and centrifuge methods, on nearly 5,000 samples from some 60 different crude oils.
The average water content and time required for the three methods were:
- Centrifuge method-0.05% water, 30 min
- Distillation method 0.200% water, 2 hr
- Karl Fischer 0.215% water, 5 min.
The test data were grouped by crude type and averaged for each type. The water amounts measured by Karl Fischer and by distillation are shown in Fig. 1. The solid line in Fig. I is the 45 diagonal, which represents the same values determined by both methods.
The data fall very close to the diagonal line, indicating good agreement between the two methods. The correlation coefficient is 0.98, which indicates good accuracy.
The majority of the Karl Fischer data lie above the diagonal line, which means that, as a rule, the test measures slightly more water than does the distillation method. This probably is because, in the distillation method, it is mechanically difficult to transfer all of the water from the distillation flask and condenser into the water trap.
The tests indicated essentially the same results for a wide range of crude gravities.
Fig. 2 shows the difference in the water contents determined by the Karl Fischer and distillation methods, plotted against API gravity. The correlation coefficient for this graph is 0.39, which is fair, but the correlation slope is only 0.0013. These results indicate that gravity does not affect water, as measured by the Karl Fischer method.
The conventional Karl Fischer method sometimes exhibits interference from sulfur compounds, which tended to react the same as water. For this reason, the LOOP study also recorded sulfur content in the samples tested, to check for sulfur interference.
Fig. 3 is a plot of the difference in water content, as measured by the two methods, vs. the sulfur content of the samples. The results indicate that sulfur contents as high as 4.5 wt % had virtually no effect on the results between the two methods, as evidenced by a correlation coefficient of 0.008 and a correlation slope of 0.00018.
The lack of sulfur interference in the coulometric Karl Fischer method as opposed to the conventional titration method is thought to be caused by the extremely rapid response rate of the coulometric method. This short response time does not allow time for the slower chemical reaction between sulfur and the Karl Fischer reagent to occur before the coulometric end point is reached.
There is evidence that aldehydes used in downhole well treatment may react with some Karl Fischer reagents and lead to apparently high water contents. This may be avoided by using Karl Fischer reagents which are not sensitive to aldehydes. Knowledgeable vendors of Karl Fischer reagents can provide guidance for specific applications.
Conventional Karl Fischer titrometry utilizes the quantitative reaction of water with iodine. This is widely used as the standard method of moisture measurement because of its speed and high sensitivy.
REACTIONS
In the coulometric Karl Fischer titration, a crude oil or other hydrocarbon sample is added to a pyridine free Karl Fischer reagent (iodine ions and sulfur dioxide are the principal components). Iodine, generated electrolytically, at the anode, reacts with water in the sample:
[see formula]
Iodine is generated in direct proportion to the quantity of electricity, according to Faraday's Law. One mole of iodine reacts quantitatively with one mole of water; therefore, 1 mg, of water is equivalent to 10.71 coulombs:
[see formula]
Based on this principle, water content can be determined directly from the quantity of electricity required for electrolysis. This determination eliminates the need for adding reagent and performing tedious standardization procedures.
Portable coulometric moisture analyzers may be used in the research laboratory as well as on the tailgate of a pickup truck with no loss of accuracy.
