Surveying sour water in crude oil

The petroleum industry typically uses water cut meters to measure and monitor the volume percent (V%) of water in a liquid hydrocarbon product flowing from a well, produced from a separator, transferred in pipelines, or stored in loading tankers. There are several technologies used for such measurements with the main being dielectric measurements using radio or microwave frequency and near-infrared (NIR) measurements. Gamma ray-based instruments are also used, but they’re less common.

These measurements are important because excessive sour water carryover throughout the crude oil supply chain results in higher corrosion rates in pipelines, off-spec crude oil supplies, and more water-to-evaporation ponds at terminals. One major challenge is finding a reliable, low-maintenance, accurate analyzer to measure low (0 - 0.2 V%) water cut at the outlets of gas-oil separation plants (GOSP). In addition, the analyzer needs the following features to measure correctly:

Installing an inline, static mixer to homogenize the oil and water mixture;
Installing sample probes to provide a representative sample for the analyzer;
Installing a lab-sample, take-off point within 1 m of the analyzer probe; and
Correct sample point location to obtain a representative sample.

One such analyzer, an electronic, dielectric-constant type measurement device, uses temperature and dielectric tables to compensate for temperature-induced dielectric changes from the process and the environment. It enables the analyzer to avoid drift after its first field calibration. The water cut meter is an insertive device with no analyzer shelter, and it requires a three-sided sunshade.

The analyzer has a built-in static mixer and lab-sample, take-off point, ensuring accurate measurement and validation. The water cut meter is a tested device with more than 16 years in the field, and it has a user list including international oil companies. It’s manufactured in flexible user ranges, which cover most upstream and downstream applications.

The advantages of this water cut include minimal maintenance, low operating expenditures (OPEX) and capital expenditures (CAPEX), and suitability for harsh weather conditions. It was piloted in the outlet of a desalter at an Aramco GOSP, and was monitored for more than a year. During that time, the water cut meter’s readings were found to be accurate when compared to lab analyses, and it had high online availability (>95%) and was shown to be robust. It also demonstrated an excellent response to process changes during field tests.

Continuous monitoring

The tested, dielectric-constant type, water-in-oil monitor is a full-bore, inline and online device designed to continuously monitor water content in flowing, hydrocarbon media (oil). It’s inline because it can fit into an existing pipeline, and it’s online because it can continuously monitor and transmit values to a remote location.

Figure 1. Typical arrangement for installation

The unit is designed so the fluid mixture flows through a fixed electrode assembly (Figure 1). This enables the dielectric constant of the mixture to be measured as a variable capacitance, which in turn produces an analog or digital output. This solution utilizes smart-sensor, data-analytics technologies in lieu of commonly used measurement techniques, and leverages built-in, artificial intelligence (AI) models. The accuracy is ±0.05% in the range from 0 - 1%. The analyzer is rated for an ambient temperature of 60 °C.

Advantages and limitations

The dielectric-constant analyzer offers distinct advantages because no frequent-field calibration is required. Once the analyzer is calibrated at the commissioning period, there’s no need for repeat calibrations, so safety is improved because operations personnel aren’t exposed to H₂S gas.

It also offers easy installation with less space requirements. It’s robust and rugged construction is suitable for harsh industrial environments and hazardous area installation. Only minimal maintenance is required with no consumables. In addition, it has a faster response time, which helps realize the full potential of advanced process control (APC) schemes. However, the meter is unable to measure water cut in multiphase stream compositions.

Field test

Saudi Aramco’s field test was conducted by creating upsets in the process and observing the response of the analyzer. Lab samples were taken before and after creating the upsets to compare with analyzer readings. The lab method used was ASTM D4007 (centrifuge method).

Test criteria

The pilot success was measured by the following criteria:

The water cut monitor reading value must be within ±0.05% water of the actual value determined by lab-sample analysis;
The monitor must operate successfully within the upper environmental temperature limits listed in the monitor manufacturer’s specifications during summer’s ambient air temperatures and radiant heat conditions;
The monitor must not show physical degradation or probe damage related to normal or abnormal operating conditions (provided the abnormal operating conditions don’t exceed the design conditions of the monitor);
The monitor must not show signs of signal or measurement interference due to process background changes that may occur during normal or abnormal operating conditions (provided the abnormal operating conditions don’t exceed the design conditions of the monitor); and
The monitor must operate reliably during the pilot period upon commissioning and must attain a 95% or greater online availability under normal and abnormal operating conditions (provided the abnormal operating conditions don’t exceed the design conditions of the monitor).

Results

The analyzer was calibrated using lab sample readings. Table 1 shows the analyzer water cut reading compared to the lab on different dates during the pilot period. The readings in Table 1 were obtained during several tests conducted by varying the water content of the process.

Table 1: Analyzer/Lab Comparison Readings

It was observed that the water cut readings agreed with the lab for high readings at 0.8%. However, the reading didn’t agree sometimes in the range from 0 - 0.2 V%, which is explained by the fact that the ASTM D4007 (centrifuge) lab method has an accuracy ±0.1%. As stated previously, the analyzer has an accuracy of ±0.05% in a water cut range from 0 - 1%.

Response to changes in the process

The water cut meter has demonstrated an excellent response to process changes during the field tests as shown in Figure 2. For example, water cut tests were conducted.

Figure 2: Analyzer response to the process changes (blue trend)

The following changes, including timing, were implemented:

9:33 a.m.—recovery pumps started and two high water wells were introduced;
9:41 a.m.—wash water increased from 1.5-1.7%
11:33 a.m.—recovery pumps switched offline

Selected analyzer readings in table format are:

9:30 AM: 0.001 V%
10:23 AM: 0.22 V%
10:50 AM: 0.38 V%
11:20 AM: 0.53 V%
12:42 PM: 0.1 V%

Conclusion

The analyzer described in this article can be used for various water cut applications in flexible user ranges that cover most upstream and downstream applications.