ISO 3171 Petroleum liquids - Automatic pipeline sampling
6 Profile testing
6.1 Introduction
This clause describes methods which may be used to test the uniformity of water distribution across the pipeline at the chosen sampling location.

6.2 Principle
Profile testing tan verify that a representative sample is available at the proposed or existing sampling location. The extent of stratification or non-uniformity of water concentration can be determined by taking samples simultaneously at several points across the diameter of the pipe and analysing these for water content. A multipoint probe, as opposed to a traversing probe, is required to eliminate uncertainties due to non-simultaneous sampling.

In order to make a valid judgement regarding uniform dispersion, a minimum of five profile tests should be made, although only three profile tests are required to demonstrate stratification. For acceptable results when using manual methods of test, the mean water concentration should be between 1 and 5 %.

NOTE - If centrifuge tubes are being used for the water determination and the water concentration is above 20 % extreme care should be taken when taking the readings because of the larger volumes between graduations.

If an automatic continuous method of determining water content is employed, then the water content should be within the working range of the instrument. To cope with conditions of complete water separation, the instrument should tolerate 100 % water without failure.

6.3 Methods
The choice is between three methods of test, depending upon the level of water concentration that tan be assured throughout the period of testing and the ability to achieve reliable and stable conditions. Generally, a level of at least 1 % water is necessary for a valid test.

- Method 1 employs the injection of additional water into the pipeline. lt is the preferred procedure and should be used whenever there is doubt that the natura1 level of water concentration in the crude oil will be consistently more than 1 % during the test period.

- Method 2 relies upon the natura1 level of water concentration. To be valid, this method should only be used when there iscertainty that the natura1 water concentration in the crude oil during the period of test will be at least 1 %.

- Method 3 is an alternative method for use with ships. Under some circumstances additional water may be introduced into the ship's compartments or storage tanks to increase the level of water concentration to at least 1 %.

NOTE - A water concentration of 1 % is enough to determine whether the dynamics of the system tan provide adequate mixing while minimizing the effects of the inherent measurement error.

6.4 Determining the water concentration profile in a pipeline and validation of the sampling location
6.4.1 Equipment
6.4.1.1 Multipoint sampling probe
The test is performed using the multipoint probe illustrated in figure 4. The probe openings should face upstream into the direction of flow, and are internally reamed to minimize entry flow disturbances. The probe should be fitted horizontally into a vertical Pipe, or vertically into a horizontal pipe. When the probe is fitted vertically, the recommended direction of flow in the probes is vertically downward where possible. In horizontal pipes, an extra straight-tube without bend is positioned vertically with the entry opening as near as practicable to the bottom wall of the pipeline (see figure 5). This Point is used to check that no free water is running along the bottom of the pipe.

The centre of the tube opening of the first and last points of sampling should be iocated 20 mm from the wall of the pipeline. All tubes leading to the external valves should be approximately the same length to ensure that simultaneous samples are taken.

A probe with a minimum of 5 sample points is recommended for 300 mm (12 in) pipe size or larger. Below 300 mm (12 in) pipe size, three sample points should be used.

NOTE - The internal bore of the tubes should be large enough to avoid blockages. For crude oil, 6 mm is the recommended minimum internal diameter.

As a safety precaution, the probe should be installed and removed during low pressure conditions. However, the probe should be equipped with safety chains and stops to prevent blow out should it be necessary to remove it during operating conditions. (See also clause 13.)

6.4.1.2 Measurement
The measurement of water content at each point across the pipeline may be made by either manual or automatic means.

When measurement is made by a manual laboratory technique, other than the centrifuge method, the sample should be collected into a receiver of adequate capacity for the test, following the procedure described in 6.4.2. When the centrifuge method is used, the Sample may be collected directly into the centrifuge tube, but it may be difficult to dispense the exact volume.

Alternatively, when continuous monitors are employed (for example capacitance cells), the samples are passed directly through individual measurement cells, one for each profile point. This procedure has obvious advantages in that profile changes tan be seen as they occur by observation of suitable indicating or recording apparatus. This automatic method also avoids the necessity for a very large number of repetitive laboratory tests.

With both methods, provision should be made for the collection and disposal of waste oil passed through the probes during the test period.

6.4.2 Operating procedure
If the application requires a range of crude oils to be sampled, the worst case conditions shall be selected. If this is not possible the formulae in annex A tan be used to extrapolate the results to other conditions. The results of extrapolation should be treated with great care.

6.4.2.1 Manual measurement
a) The sampling frequency should provide at least one profile every 2 min and at least five profiles at the minimum flow rate with a water concentration of at least 1 % (m/m).
b) If method 1 (ses 6.3) is used, water shall be injected upstream from the multipoint probe, at a sufficient distance from the probe and upstream of any mixing elements, so that a representative profile is obtained, but in a manner which will not create an energy dissipation rate significantly higher than the existing one [see annex A, A.4.2.2, equation (26)].
c) The injection operation should be made in accordance with figure 6, using a pump that will produce the proper flow rate and pressure. If necessary, water should be injected in successive percentages of 1, 2, 3, 4 and 5 % or more of the volume of crude oil flowing in the pipeline. The operation should be repeated at five representative but different flow rates and particularly at the minimum flow rate. These should be constant to within 10 % during the period of each test.
d) If method 3 (sec 6.3) is used, the ship's compartment to be spiked with water should be selected such that it tan be pumped without creating trim problems.
e) Each element of the probe should be rinsed out thoroughly before sampling by running through it an adequate amount of crude oil.
f) Flow rates in the multipoint probe should be regulated before sampling so that the entry velocity at each point of the probe is the same and is preferably equal to the velocity of the liquid flowing in the Pipeline. If not, any differentes should be recorded.
g) Sampling should begin before the calculated water arrival time and continue, without alteration of the flow rates in the probe, until sufficient profiles have been taken to obtain a minimum of five which meet the requirements of 6.2.
h) Sampling should continue during the time-period that injected water is passing. The time-lag between water injection and arrival at the sampling location should be taken into account (see figure 6).

6.4.2.2 Automatic measurement
a) The provisions of 6.4.2.1 b), c), d), e), f), g) and h) apply.
b) A recent development is the use of capacitance cell profile test systems to demonstrate continuously the degree of homogeneity across the profile and with time. The capacitance of water in crude oil mixtures will vary with crude oil composition; it will also vary with temperature and may vary with water droplet size and shape. The individual cells comprising such an automatic profiling system should be zeroed on the particular crude oil being sampled, and steps taken to ensure that temperature and flow rate differences between cells are minimized. In profile testing, the measurement of interest is the relative concentration of water at each point across the pipe. Therefore, provided that capacitance cells with similar characteristics are used, there may be no need to adjust the span for different crude oil types.
c) Timing is much less critical than for manual testing, as is the need for steady conditions. A visual display will indicate the profile conditions as they occur, enabling the worst (or best) conditions to be identified.
d) Figure 7 shows a typical bar graph presentation, based on experimental work, from a profile test at three different sampling locations along the pipe.

6.4.3 Typical test data
6.4.3.1 Manual measurement
Annex B Shows some typical test data from a profile test at a crude oil terminal. Tests were made at two flow rates and six profiles were obtained for eight points across the pipe for each flow rate.

After excluding all profiles where the water content is less than 1 % (m/m), the Point averages, profile averages and overall mean are calculated. The next step is to obtain the percentage deviation of each point average from the overall mean. profiles should be drawn as shown in annex B.

6.4.3.2 Automatic continuous measurement
With automatic profile test equipment, using continuous water monitors connected to a microprocessor, it is possible to continuously display and record the water profile in any suitable form such that changes caused by flow rate variations or injected water percentage tan be observed as they occur. The effect of variable energy mixing devices tan also be observed.

Similarly, calculation of mean and average water contents and percentage deviations at points across the pipe tan be carried out instantaneously. These results tan be displayed or recorded with the profile and other test data.

6.4.4 Presentation and interpretation of results
Each profile may have the general appearance of one of the types P0 to P5, shown in figure 8.
a) P0 represents the most suitable distribution, since the concentration ratio is within the acceptable accuracy limits at any point of the cross-section.
b) P1, and P2 represent acceptable distributions, provided that the sampling probe is positioned according to the recommendations in 4.4.2 and 5.4.1.
c) P3 to P5 represent unacceptable distributions with concentration ratios outside the acceptable accuracy limits. Any profile beyond the acceptable accuracy limits indicates that sampling results cannot be expected to be representative. To achieve an acceptable distribution either a higher flow rate or a mixing device should be used.

Annex A gives a simplified theoretical relationship between flow rate and oil characteristics on the one hand, and the profile on the other hand.