Crude Oil Quality Association
Water Measurement in Crude Oil
Discussion Paper
Updated - October 2, 2003
INTRODUCTION
The Crude Oil Quality Association established a subcommittee of members who are experts in the field of water measurement in crude oil. The purpose of the subcommittee was to publish a paper outlining the various methods available to measure water in crude oil and to provide an overview of some industrial experiences on the measurement process. General statements on the advantages and disadvantages of each measurement type are also discussed.
The basic measurement techniques for water in crude oil are:
Method Test # Comments
Centrifuge (Lab) ASTM D4007 water and sediment results
Centrifuge (Field) ASTM D96 water and sediment
Centrifuge (Field) API Chapter 10.4 water and sediment
Distillation ASTM D4006 water results only
Karl Fischer ASTM D4928 water results only
Online Water Devices API draft standard is in the API publication department
The following observations apply to the accuracy of water measurement methods:
- Procedure: It is critical that the approved, written procedures be followed for all test methods.
- Methodology: There are many variations of these tests that apply to substances other than crude oil. These variations should not be used to measure water in crude oil, as the results may not be correct.
- Reagents and Solvents: The use of proper Karl Fischer Reagent is important. Premixed solutions are commercially available or 60 parts reagent can be mixed with 40 parts xylenes for better solubility. Some manufacturers may use a different reagent mix even though other ratios are not within the API/ASTM published Standards/Test Methods. The proper use of solvent in distillation and centrifuge methods is equally important.
- Audit: Auditing of procedures must be done regularly, in lab and field. Preparation of the reagents and the handling of them are crucial and should be included in the auditing process.
- Sampling: Sampling procedures are critical to accurate water measurements, regardless of the water test method used.
CENTRIFUGE
It is generally accepted that the centrifuge methods, both the lab (D4007) and the field (D96), are not the most accurate. Disparities between the lab and field methods can occur due to the differing treatment of the solvent used. The centrifuge can only be validated by adding a known volume of water to the sample and rerunning the test to verify the procedure. The Karl Fischer and distillation test methods are easily validated with a quality control sample.
The centrifuge method does generally not pick up emulsified or dissolved water. Using a de-emulsifying agent before the test is run (as is noted in the written procedures for the method) can increase the chance of the centrifuge reading more of the water content. Aging the sample for several days and allowing the water to settle out will also increase the chances of the centrifuge method detecting more of the water. However, the additional time spent aging the crude sample is not advantageous for normal operating time frames.
The inaccuracies of the centrifuge are well documented. In fact the IP (the Institute of Petroleum of the United Kingdom) has eliminated the centrifuge method from its set of standards. It should be noted, however, that centrifuge was not in common use in the UK in recent years. Also, ASTM has withdrawn D96 from its standards. It is important to note however, that in the US, the centrifuge method will most likely keep its place in the industry because of speed, prevalence and ease. Additionally, the method yields results for both sediment and water, an important consideration in domestic crude production where sediment can vary from batch to batch. API (API Manual of Petroleum Measurement Standards, Chapter 10.4) does continue to maintain an equivalent centrifuge method in its manual at present.
DISTILLATION (ASTM D4006)
Water measurement by distillation is a laboratory procedure. Pipelines or refineries rarely use ASTM D4006 for day to day operations. The major use of D4006 for refineries and pipelines is to resolve a dispute when a number cannot be agreed upon between delivering and receiving parties. Certain customs agencies require D4006, so it is more commonly used for imported crudes. ASTM D4006 does not provide sediment results.
KARL FISCHER (ASTM D4928)
Karl Fischer titrimetry is a more accurate moisture measurement method utilizing the quantitative reaction of water with iodine. The method is becoming recognized as the standard method of water measurement because of its speed and precision. According to API/ASTM published standards; Karl Fischer is both more precise and faster than either centrifuge or distillation. Please see Attachment 1 (at the end of this paper).
Similar to distillation, Karl Fischer results do not include sediment. A separate test would be necessary to determine the amount of sediment in a sample.
Although the equipment necessary to run a Karl Fischer is not, at $5000 per instrument, individually high priced, it is not prevalent. There is maintenance involved as well; the solutions must be maintained properly and changed on a frequent basis. QA/QC operations must be performed to ensure validity of the results. At present, Karl Fischer is primarily a laboratory procedure. Although a portable Karl Fischer instrument can be truck mounted, there are inherent handling problems related to the glass components of the equipment. Also, maintaining the cell in a manner that precludes intrusion of water from the atmosphere or other sources can create difficulties.
There have been some studies that indicate Karl Fisher may be susceptible to interference from naturally occurring species in the crude oil. For example, crudes with high mercaptans tend to be prone to the interference. A methodology to detect the extent of the possible interference in a particular crude will be presented to ASTM shortly.
ONLINE WATER DEVICES
The API has written a draft standard on Online Water Measurement Devices (OWD). The draft standard is in publication.
There are currently at least 25 different models of online measurement available. The technology for all the manufacturers falls into one of three categories:
Microwave: generally considered the most consistent
needs to be corrected for temperature and density
Capacitor: oldest technology, still evolving
needs to be corrected for temperature and density
Optical: many different wave lengths, different light sources
up and coming technology
consistent
Some COQA member companies (Arco of California, Enbridge Pipeline) use online devices with variable successes.
SEDIMENT
There are three main ways of measuring sediment in crude oil:
- Extraction ASTM D473
- Filtration ASTM D4807
- Centrifuge API Manual of Petroleum Measurement Standards, Chapter 10.4
Methods to measure sediment quickly are being researched but are not commercially available yet. When these methods become available, it is expected that there will be industry acceptance of separate tests for sediment and water.
When sediment levels are significant, accurate measurement is important for all aspects of the petroleum industry. It is also critical to properly measure sediment when the level varies significantly from batch to batch
It was noted that foreign crude cargoes brought into the United States sometimes use the Karl Fischer method for water determination. Sediment by centrifuge is generally performed for these crude vessels. When that result indicates a possible problem, extraction or filtration may be performed to pinpoint the sediment number. Sometimes, parties agree on a sediment number for each crude and it is used for a set period of time, such as thirty days. After the agreed time passes, a new sediment number is negotiated. Although that methodology has worked well for foreign crudes, it is perceived to have drawbacks for US lease production where sediment levels vary more frequently.
In US domestic production, the oil is not "settled" into large tanks but is batched through with little retention time into the pipeline system. The batches can vary considerably in sediment content, making a fair, negotiated estimate difficult to ascertain. Sediment at US lease production is unstable and could be significant.
Sediment as it moves through the transportation system in the US is generally not an issue, unless the amount is significant and operations are affected. Sediment received at the refinery, however, can cause serious problems. The material causes fouling throughout the distillation process and can impact product quality.
SUMMARY/CONCLUSIONS
The importance of water measurement is not disputed. In fact, many companies have observed a trend of increasing water in crude oil. The use of chemicals to enhance production or restart wells is a possible cause of higher water. While these phenomenon may be temporary, they occur often enough to emphasize the necessity of consistent water measurement.
Water in crude oil must be treated before it can be discharged into a receiving body of water. All oil industry locations operate under stringent discharge permits. The costs of treating water to a dischargeable level are expensive. Therefore, accurate water measurement is crucial to determine the proper economics of a crude oil.
There are economic ramifications to both the pipelines transporting the water and to the refineries receiving it. Pipelines incur operational costs to move water even though those costs are not necessarily reimbursable. One of the COQA member companies estimated that in its 1MMBPD refining system, over 400,000 barrels of material were annually unaccounted for due to centrifuge readings versus Karl Fischer combined with sediment by extraction numbers. Please see Attachment 2.
ATTACHMENT 2A - VESSEL CHART Document
ATTACHMENT 2B - REFINERY CHART Document
A Karl Fisher test method is also the method of choice when establishing a refining value for a particular crude. The difference between the Karl Fischer reading and the centrifuge reading actually utilized in the transaction has economic ramifications. Although companies establish and utilize refining values in different ways, there is consensus that water measurement with Karl Fischer accuracy is critical in arriving at the proper economics.
Water measurement accuracy is open to interpretation in the context of handling the crude. A custody transfer is different than refining operations. The observation was made that Karl Fischer tends to yield higher water-only results than a centrifuge water-and-sediment number. Perhaps Karl Fisher could be used to obtain the most accurate water measurement with sediment handled separately for refining operations and those occasions where the sediment could be high. The Karl Fischer / centrifuge correlation is, however, very crude specific. The crude itself and the way it is handled (the adding of de-emulsifying agents at the production location, for example) definitely affect the test results.
The Karl Fischer method appears to have overwhelming advantages with regards to speed and precision when compared to the other, currently available water measurements for crude oil. The COQA, dedicated to quality, endorses efforts on the part of the industry to use Karl Fischer, noting the disadvantages which include costs and lack of a sediment determination. It is not expected that, with these drawbacks, the US oil industry will exclusively use Karl Fischer in the near future. However, individually establishing a bias between centrifuge and Karl Fischer for each crude can help in the interim with company specific economic and operational decisions. It may be feasible to use that same relationship, on a macro level, to show economic incentive for a commitment to Karl Fischer equipment.
A partial listing of technical references can be found in Appendix A.
Appendix A
Attachment 1:
Test Method | Time Required for Test | Accuracy @ 0.1% Water |
Centrifuge Method | | Reproducibility |
ASTM D4007 | 30 minutes | 0.2 |
Distillation | | |
ASTM D4006 | 120 minutes | 0.11 |
Karl Fischer | | |
ASTM D 4928 | 10 minutes | 0.02 |
* Time estimate is for single analysis after stabilization of equipment. The equipment must also be cleaned at a frequency based on the crude types analyzed and the number of samples analyzed.