Water
Characterization Methods
BP has provided, the PWRI Consortium with some of their guidelines for determining relevant properties of injection water. This document is extracted largely intact. The original reference is “Water Characterisation Methods Used by BP,” by Ian McCracken.
Introduction
Methods have been developed for assessing the characteristics of seawater or produced water to its injection or re-injection to identify the level of suspended particles present and the nature of these particles. A series of tests has been developed, some based upon standard test methods, and others developed for specific needs, in order to provide the required data. Your own organization may have adopted slightly different protocols, but the procedures described here are a relevant reference.
The following test methods are described individually below:
1. Total Suspended Solids
2. Oil Content
3. Dissolved Oxygen Content
4. Water Chemistry
5. Bacterial Analysis
6. Solids Examination
In each case, important aspects of the sampling procedures will be discussed in order that analyses may be made on representative water samples. In all cases, it is vital that clean sampling containers be used, and that all sample lines are thoroughly flushed prior to sample collection. To be of greatest value, produced water destined for re-injection should be sampled as near to the point where the fluid first hits the formation around the injection well as is possible.
1. Total Suspended Solids Determination
|
Background: |
This technique relies upon filtering a known volume of water and weighing the amount of filtrate remaining on the filter paper, after removing any oil. |
|
|
|
|
Equipment: |
Standard filtration equipment |
|
|
Whatman GF/F Filter paper |
|
|
Balance |
|
|
|
|
Sampling: |
Care should be taken to ensure
that the water is not allowed to generate more particles after sampling. This can occur usually through either
mineral scale formation or through iron flocculation. An indication of the tendency to both
can be obtained from water chemistry determinations (see later). |
|
|
|
|
|
Upon a pressure drop, dissolved
carbon dioxide may be released. This
raises the pH and increases any tendency to deposit calcium carbonate. Hence it is recommended that samples
are collected in a pressurised bomb.
As an alternative, acidifying the sample with a few ml of 10% HCl will
also prevent further calcium carbonate precipitation. |
|
|
|
|
|
Exposure to air will oxidise any
dissolved iron II, forming insoluble iron oxides and hydroxides. These may precipitate, increasing the
apparent solids loading. It is
very difficult to eliminate the chances of contact with oxygen, even when
collecting in a pressurised bomb.
Hence we recommend sampling into a reducing acid, such as erythorbic
or citric acids, which will tend to keep the dissolved iron in solution. A concentration of 100ppm is usually
more than adequate providing analysis is carried out within an hour or two of
sampling. |
|
|
|
|
|
If there is any doubt over
whether particle generation is occurring, we recommend taking several samples
and analysing them over a period of hours. Any increase in TSS with time implies particles are being
produced. |
|
|
|
|
Technique: |
1. Weigh a Whatman filter, and place
in an oven at around 50° C for at least 2 hours before use. |
|
|
2. Filter a known volume (typically 500ml
to 1litre) of sample water. |
|
|
3. Wash the residue remaining on the
filter paper with |
|
|
4. Wash the residue with filtered
deionised water to remove dissolved salts. |
|
|
5. Oven dry the filter paper and
re-weigh. |
|
|
6. Results quoted in weight of solid
per volume of sample (usually mg/l). |
2. Oil Content
|
Background: |
This technique relies upon
extracting the oil from the collected brine sample, analysing it using a UV/Visible
spectrophotometer and comparing the absorbance against a set of standards. |
|
|
|
|
Equipment: |
UV/Visible Spectrophotometer. |
|
|
|
|
Sampling: |
Care should be taken to ensure
that the container surface is not oil wetting. |
|
|
|
|
|
While oil flocculation is a
potential problem if the oil droplet particle size distribution is to be
determined, the particle size is irrelevant to the total oil content
determination. |
|
|
|
|
Technique: |
1. Examine
a series of standards of known oil content and construct a calibration graph
of oil content against absorbance at the required wavelength. |
|
|
2. Acidify
a known volume (about 100ml, depending upon the ultimate oil content) with a
few drops of 10%HCl. |
|
|
3. Extract
the oil with an equivalent volume of |
|
|
4. Filter
the extract through a Whatman GF/F filter. |
|
|
5. Analyse
the filtered extract at the required wavelength in the spectrophotometer. |
|
|
6. Compare
the resultant absorbance against standards. |
3. Dissolved Oxygen Content
|
Background: |
There are several commercial
oxygen probes available which produce a direct reading of dissolved
oxygen. Whichever instrument is
selected should be capable of measurement to ppb levels. Ideally, oxygen content
determinations should be conducted 'in-line' to avoid inevitable exposure to
oxygen when taking samples. |
|
|
|
|
|
As an alternative, a chemetrics
oxygen test kit could be used, but these tend not to be sufficiently
sensitive, and suffer from the need to take a fluid sample. This test relies upon exposing the sample
to a reagent, which reacts to oxygen by changing colour. Comparing the colour to a set of
standards provided with the kit provides the correct oxygen level. However, without immediate analysis,
exposure of the sample to air renders the determination invalid. |
|
|
|
|
Equipment: |
We have successfully used an
Orbisphere 'in-line' oxygen meter. This can measure to 0.3ppb and is flow independent
within a broad range of flow rates and temperatures. |
|
|
|
|
|
The Chemetrics test kit is available
through CHEMetrics Inc, Calverton, Virginia 22016. |
|
|
|
|
Sampling: |
The sample point should be
flushed well to remove any air bubbles.
For the Orbisphere oxygen meter, sample flow should be reduced to about
100ml/min and diverted into the in-line cell for measurement. |
|
|
|
|
|
A small sample tube is provided
with the Chemetrics test kit, which should be thoroughly purged to remove air
bubbles before taking a sample. |
|
|
|
|
Techniques: |
The Orbisphere meter provides a
direct read out of dissolved oxygen |
|
|
|
|
|
The Chemetrics test kit method
involves snapping the end of a small glass ampoule containing the reagent in
the sample fluid. The sample
mixes with the reagent, causing a change in colour proportional to the level
of oxygen present. Direct
comparison with standards provided enables the dissolved oxygen content to be
determined. |
4. Water Chemistry
|
Background: |
A variety of information can be inferred
from analysing the chemistry of a sample of brine. For instance, the ion composition will provide an
indication of scaling potential, or the likelihood of iron flocculation and
oxidation. Inevitably, for a
full analysis, samples must be taken and transported away from the production
or injection site to a suitable laboratory. This means that samples may change considerably from the
time that they were taken to the time when they are analysed. It is vital to treat them correctly
in order to obtain good data.
Unfortunately, a variety of different 'treatments' may be necessary to
ensure that all data is valid. |
|
|
|
|
|
The type of data that we have
obtained include, ion concentrations for Na, K, Mg, Ca, Sr, Ba, Fe(II),
Fe(III), Mn, Cl, SO4, HCO3, other elements such as Si, S, P, B, and pH,
density etc. |
|
|
|
|
Sampling: |
There are many possible changes,
which may take place to produced water after sampling which can affect the
varied information described above.
It is frequently necessary, therefore, to take more than one sample
and treat the samples in a different manner. For instance: |
|
|
|
|
|
A steel container cannot be used
for Fe or Mn determinations - a suitable polyethylene screw cap container is preferable. |
|
|
|
|
|
Dissolved calcium ions may be
lost from solution as calcium carbonate scale - sample into a known volume of
acid to maintain calcite solubility. |
|
|
|
|
|
Dissolved Fe may be oxidised and
lost from solution as rust - sample into a known volume of a reducing acid
such as erythorbic or citric acid. |
|
|
|
|
|
pH and bicarbonate ion
concentrations can vary with time as dissolved gasses are slowly released -
these analyses should be carried out immediately after sampling, as close to
the sample point as possible.
Alternatively, pressurised samples may be taken, preserving them for
subsequent analysis, which again should be carried out immediately after
depressurising. |
|
|
|
|
The general rule is to ensure that whatever analysis is to
be undertaken at a remote laboratory, the sample cannot change from the
moment that the sample was taken. |
|
|
|
|
|
Techniques: |
(i) pH : We recommend a suitable
portable pH probe with a temperature adjustment in order to achieve an
immediate readout of pH. |
|
|
|
|
|
(ii) Bicarbonate Ion
Concentration: A suitable
acid/base titration method is recommended since it is often possible to set
up at the sample point. |
|
|
|
|
|
(iii) Chloride Ion
Concentration: The Mohr titration is recommended since it is quick and
accurate. This has been used to
identify ionic strength differences in samples taken over a period of a few
days, and can highlight when to expect changes in the concentration of other
ions. |
|
|
|
|
|
(iv) General Cation and Anion
Analysis: Atomic Adsorption spectrophotometry provides an accurate assessment
of ion concentrations. |
5. Bacterial Analysis
|
Background: |
This is one of the most
difficult analyses to perform and produce meaningful data. Frankly, in our experience there is
no one foolproof simple technique.
We have relied upon the experience of microbiologist groups who have
worked on the reservoirs we targeted for study, and taken their advice on
test methods. The precise
methods do appear to vary from reservoir to reservoir, making comparative
data collection difficult. In
general, we have used techniques, which aim to identify the level of Sulphate
Reducing Bacteria (SRB's) and General Aerobic Bacteria (GAB's). |
|
|
|
|
Sampling: |
Samples should be analysed as
soon as possible after taking them.
This is vital for quantitative data since a delay could lead to
bacterial activity, apparently increasing the concentration of bacteria in
the sample. |
|
|
|
|
Techniques: |
We have used two types of
methods: an extinction dilution technique and a gene probe. |
|
|
|
|
|
(i) Extinction Dilution
Technique: |
|
|
This relies upon growing bacteria
in a suitable culture medium using decreasing amounts of sample and
identifying the least dilution where bacteria are detected following an
incubation period. The precise
culture medium depends upon the bacterial types, and varies for SRB's or GAB's. |
|
|
|
|
|
Collected samples
should stand for a few minutes in order to allow oil to separate from the
brine. Using a sterile syringe
and needle, 1ml of brine is injected into the first test kit bottle
containing the growth medium. After
shaking this bottle, 1ml is abstracted using a clean, sterile needle and
syringe, and injected into the second bottle containing the growth
medium. This is repeated until
the sixth bottle, which therefore contains a sample with a 1:100000 dilution
of the first sample. After 28
days incubation (at a constant temperature of 35 Deg C) each bottle is
examined for bacterial growth.
Quantitative information is obtained by identifying the least dilution
at which bacterial growth is found. |
|
|
|
|
|
Growth media and further
information may be obtained from Oilfield Microbiological Services (0224
249424) or SGS Redwood (0224 770668), both of Aberdeen. |
|
|
|
|
|
(ii) Gene Probe and
Rapidcheck II: |
|
|
These techniques rely upon
examining for the genetic makeup of the bacteria. The gene probe was developed specifically for our Prudhoe
Bay field; DNA from detected bacteria providing a direct measure for both SRB
and GAB bacteria. |
|
|
|
|
|
Rapidcheck is a similar technique
for SRB determinations developed by Conoco, which is now being more widely
used in the industry. |
|
|
|
6. Solids Examination
|
Background: |
BP has employed a number of
methods to characterise the nature of the solids found in sea water for
injection or produced water for re-injection purposes. It is valuable to know whether the
solids are corrosion products, mineral scale, oil, sand etc. Such knowledge could be used to
identify a need for an improved chemical treatment, better sand control or
alternative filtration methods.
In our case, we were also interested in identifying typical particles
and their size distribution to define experimental parameters and for
modelling purposes. |
|
|
|
|
Sampling: |
As discussed previously,
sampling must be fit for purpose.
For instance, collecting solids on a filter paper for subsequent
analysis in a remote laboratory would be a waste of time if the solids were
predominantly iron products, which could oxidise quickly. |
|
|
|
|
|
We recommend using a simple
'in-line' filtration apparatus, which can take a wide range of pore size
nucleopore filters. Washing the
filter paper to remove oil (using |
|
|
|
|
Technique: |
The techniques used to examine
collected solids were as follows: |
|
|
(i) Electron Microscopy with
Energy Dispersive Spectrometry.
This provided both a visual image and identification of the elements
present in any selected particle. |
|
|
|
|
|
(ii) X-Ray Diffraction. This technique enables all
crystalline material present to be identified. |
|
|
|
|
|
(iii) Coulter Counter Particle
Size Distribution. This
technique provided number size distributions, volume distributions, counts
per volume of sample and the mean particle size. A technique was developed to determine data for solids +
oil, solids only and oil only.
Solids + oil were determined directly, solids only were determined by
removing the oil by washing with |
|
|
|
Modified: 6/24/02
Back
to Key Issues