SPE 87470
How Scale Inhibitors Work: Mechanisms of Selected Barium Sulphate Scale inhibitors
Across a Wide Temperature Range
K. S. Sorbie1 and N. Laing2
1 2
Heriot-Watt University, Edinburgh, Scotland; BP Exploration, Aberdeen, Scotland
Copyright 2004, Society of Petroleum Engineers Inc.
th
This paper was prepared for presentation at the 6 International Symposium on Oilfield Scale
held in Aberdeen, UK, 26-27 May 2004.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
Abstract
Scale inhibitors (SI) have been applied very successfully over
many years in oilfields to prevent the formation of mineral
scale. Both barium sulphate and calcium carbonate scales may
be prevented using inhibitors, although in this work we will
focus on the more difficult barite inhibition problem. A
number of publications have appeared discussing the
mechanisms by which barium sulphate scale inhibitors operate
to prevent or retard scale formation. The mechanisms are
discussed here in terms of (a) nucleation inhibition where the
scale proto-crystals forms but are then disrupted or redissolved
by the action of the inhibitor molecules, and (b) crystal growth
inhibition where the inhibitor is thought to adsorb or interact
with the active crystal growth sites (growing edges or spirals)
hence retarding or stopping the crystal growth process. Both
of these mechanisms are consistent with the inhibition of
mineral scale at “threshold” levels, typically for MIC = 0.5 –
20 ppm (MIC = minimum inhibitor concentration for a defined
level of inhibition for a given test procedure). The MIC is
always considerably below stoichiometric values in terms of
the scale inhibitor to mineral scale molar ratios. It is known
that most inhibitor types from the small molecular
phosphonates (e.g. DETPMP) to polymeric species (e.g. PAA,
PVS, PPCA) actually operate through both of the above
mechanisms although one of these may predominate for
specific species. Previous work has established that, broadly
speaking, smaller phosphonates operate principally as crystal
growth inhibitors and polymeric species work mainly as
nucleation inhibitors.
A number of factors are known to affect the inhibition
efficiency (IE) of scale inhibitors against barite formation, the
main ones of concern here being pH, temperature and the
calcium and magnesium levels in the scaling brine mixture.
The general observed effects of these factors have been
described in the literature and will be discussed in detail in this
paper. However, no complete description of the mechanism of
barite inhibition has appeared which clearly and consistently
explains all of the observed effects of these parameters for
different scale inhibitor types. It is the central aim of this
paper to present a complete and consistent set of mechanisms
for barite inhibition which may vary in degree for different
inhibitor types. Our proposed mechanisms are based on a
wide range of observations from the open literature analysed
with our own experimental and modelling results.
Experimental Details
To develop a set of mechanisms for barite inhibition, we use a
number of experimental techniques and data sources. Since
the experimental techniques used are well known, we refer the
reader to literature descriptions for details.
Static inhibition (bottle) tests and dynamic (tube blocking)
tests are both very well known techniques for determining
inhibition efficiency (IE) and are described in detail
elsewhere1,2. The brine compositions used in the new static
IE results presented here are given in Table 1. The scale
inhibitors used in this study are: diethylene triamine penta
(methylene phosphonic acid) (DETPMP), phosphino
polycarboxylic acid (PPCA) and polyvinyl sulphonate (PVS)
and their structures are given in Fig.1
Crystal structure measurements of the a-axis deformation
of the barite lattice from this work and the literature are used
here3, 4. The theory behind this is given in Ref. 5.
We also present some simple calculations of the
equilibrium system containing calcium ions, magnesium ions
and scale inhibitor. Results are given for the Ca/Mg/DETPMP
system close to MIC levels for which the stability constants
are known6.
2 K.S. Sorbie and N. Laing
Analysis of Results
To make a complete analysis of both known literature and our
present results, we consider each parameter in turn, in order
pH, temperature (T) and then the calcium/magnesium content
([Ca2+] and [Mg2+]) of the brine mixture.
Results 1 – pH
It is generally known that most common scale inhibitors will
not function (i.e. inhibit scale effectively) below a certain pH
level. The reason for this can be seen by considering the
dissociation of a generic multi-protic scale inhibitor molecule
HnA as follows:
Hn A ⎯⎯
K1
←⎯ ⎯→ H + + H n −1 A−
H n −1 A− ⎯⎯
K2
←⎯ → H + + H n − 2 A2 −
⎯
(1)
....
⎯⎯⎯
K x-1
H n − x A x − ←⎯⎯ ⎯→ H + + H n − x −1 A( x +1)−
A range of pKa values (pKa1 = -log K1, etc.) is associated with
each of the dissociation in the above equations e.g. for
DETPMP (H10A), pKa1 and pKa2 estimated at 0.5 and 1.6, with
pKa3 ~ 2.8 up to pKa10 ~ 12 with a “mean” pKa ~4.5 6. For
PVS, the minimum pKa1~ 0.5 – 1.0 (estimate) to much higher
values since it is a polyelectrolyte (mean pKa ~ 3). For
DETPMP, modeled as H10A, the speciation as a function of
pH can be calculated and is shown in Fig. 2. At pH 4 for
example, there are two main Hn-xAx- species with x = 3 and 4.
A similar speciation diagram can be constructed for any
inhibitor poly-acid for which the pKi values are known, i =
1,2, …n. The conclusion from this is that it is only the
dissociated entity Hn-xAx- that inhibits, not the neutral
protonated form of H10A. This explains why PVS (mean pKa ~
3) will inhibit scale down to a much lower pH than DETPMP
(mean pKa ~ 4.5) as shown in Fig. 3 4. The results in Fig. 3
further indicate that, for DETPMP, the species involved in the
inhibition of scale require a dissociation level with Hn-xAx-
having x ≈ 5 or 6 since at lower pH values where x ≈ 1,2 or 3,
the DETPMP does not inhibit. A further important point to be
noted is that the calcium and magnesium will only bind to the
dissociated species, Hn-xAx-, and we return to this issue later.
Results 2 – Effect of Temperature
It is know that phosphonates (e.g. DETPMP) tend to have a
lower “cut off” temperature below which they are much less
effective. Inhibition efficiency (IE) results for the 3 scale
inhibitors in Fig. 1 are shown for T = 5oC, 50oC and 95oC in
Figs. 4 and 5 for the Brent/seawater (SW) and Forties/seawater
brine mixtures, respectively (compositions in Table 1). The
Brent/SW mixture is inhibited at scale inhibitor
concentrations, [SI] = range 0.1 to 3 ppm (see Fig. 4) and the
Forties/SW mixture levels are, [SI] = 10, 20 and 30 ppm since
the Brent system has a much lower scaling tendency than the
Forties mixture. Fig. 4 (Brent) shows that, at the highest level
used for DETPMP (3 ppm), good IE is only achieved at 95oC.
SPE 87470
At 50oC and below, 3 ppm DETPMP shows very poor IE
which is indeed below the efficiency of the 1 ppm DETPMP at
95oC (Fig. 4). In contrast, 1 ppm of PVS efficiently inhibits
the Brent/SW mixture at all temperatures and even works well
at 0.5 ppm (Fig. 4). The PPCA in this figure effectively shows
“intermediate” behaviour between that of the DETPMP and
the PVS in that the IE is better at higher temperatures but does
not drop off as sharply as that of DETPMP at lower
temperatures.
For the more severe scaling Forties/SW brine mixture, the
results in Fig. 5 show similar behaviour for the DETPMP
which now only inhibits effectively at 30 ppm and at 95oC.
The PVS results in Fig. 5 are very interesting since we now
find that all concentration levels 10 to 30 ppm of PVS inhibit
completely only at 5oC. At 50oC and above, the inhibition
level of PVS even at 30 ppm is poor; indeed, the IE of PVS is
decreasing with temperature. The PPCA results for this case
are again “intermediate”, showing some increase in inhibition
efficiency with temperature and much less of a drop off as
temperature decreases than DETPMP. We note that the
supersaturation (Sp) of the scaling mixture is much higher
(approximately x10) at 5oC than at 95oC and hence the
thermodynamic driving force is significantly higher at the
lower temperature.
In summary, we find that:
- DETPMP works better at higher temperatures (where the
supersaturation, Sp, is lower);
- PPCS follows a similar trend to DETPMP but its IE drops off
much less dramatically than DETPMP at temperature drops;
- PVS works much better at lower temperatures (where Sp is
much higher).
To examine the effects of supersaturation (Sp) in more
detail, compositions of brine were varied at different
temperatures to give modified Forties/SW mixtures which
were isosupersaturation solutions. These brines were
designed using scale prediction software and details are given
in Ref. 4. IE results for the 3 scale inhibitors at 5oC , 50oC and
95oC for iso-supersaturation ratios Sp = 310 and 3190 are
presented in Figs. 6 and 7. At Sp = 310, results in Fig. 6 show
that DETPMP fully inhibits at all concentrations used (10 – 30
ppm) and shows no particular effect of temperature. That is,
the IE of DETPMP is simply affected by the value of Sp and
not the temperature. The corresponding results for PPCA and
PVS in Fig. 6 are quite intriguing. The PVS shows a very
clear decline in performance at Sp = 310 as temperature is
increased from 5oC to 95oC. Since the thermodynamic driving
force (Sp) is constant, then this indicates a clear kinetic effect
in the mechanism through which the PVS operates. PPCA is
again showing “intermediate” behaviour where it shows
increasing IE as temperature increases (like DETPMP) but is
still has some IE at the lower temperatures (like PVS). These
findings are supported by the Sp = 3139 results in Fig. 7 where
the only successful inhibition is seen for the PVS at 5oC which
rises, as expected, over the PVS concentration range from 10
to 30 ppm. In summary, we find that:
SPE 87470 How Scale Inhibitors Work: Mechanisms of Selected Barium Sulphate Scale inhibitors Across a Wide Temperature Range
- the IE of DETPMP appears to depend entirely on the Sp of
the barium sulphate solution;
- PVS works principally through a kinetic nucleation
inhibition effect;
- PPCA shows “intermediate” behaviour.
Results 3 – Calcium and Magnesium
We now consider the effect of calcium and magnesium ion
concentrations on the IE. It has been reported previously that
the IE of DETPMP depends on the presence of calcium ions
but that it is “poisoned” by magnesium1,7,8,9. Fig. 8 shows the
IE of PPCA, PVS and DETPMP as functions of calcium
concentration, [Ca2+]. The IE of both PPCA and PVS does
increase gradually as [Ca2+] increases but the increase is much
more dramatic for DETPMP. This is supported by the more
recent results for DETPMP and PVS in Fig. 9 which shows IE
results (at 2 hours and 22 hours in the static tests) for a Forties
type/SW brine mixture. The effect of magnesium is shown in
Fig. 10 where it clearly reduced the IE of DETPMP but barely
affects the performance of the PPCA or PVS. Thus, we
conclude that:
- DETPMP will not work effective at low [Ca2+] implying that
the calcium plays an important role in the inhibition
mechanism for this SI;
- calcium has a smaller but still beneficial effect of the IE of
PPCA and PVS;
- magnesium effectively “poisons” DETPMP (lowers its IE);
- magnesium has little effect on the IE of PPCA and PVS.
Results 4 – Adsorption of Scale Inhibitors on Barite
One possible hypothesis on IE effect of calcium and
magnesium (and also T) may be related to how these cations
(or T) mediate adsorption of SI onto the growing barite
crystals. An extensive set of adsorption tests were performed
at this laboratory for a wide range of conditions (pH, T, [Ca2+]
and [Mg2+]) for DETPMP, PPCA and PVS onto barite
crystals10. This study indicated that there is really no
correlation between the measured level of SI adsorption under
a given set of conditions and the corresponding IE sensitivity.
Indeed, several trends were observed which were exactly
contrary to these trends e.g. at 95oC, PVS adsorption onto
barite is higher than at 50oC but its IE is lower. We therefore
discount simple SI adsorption onto the barite crystal as being
the central mechanism of inhibition although we will later
acknowledge that some type of SI-barite crystal interaction
does occur.
Results 5 – Barite Crystal Lattice and Calcium Inclusion
We now turn our attention to the effects of calcium and
magnesium on the barite crystal lattice which is formed under
various conditions and which has been the subject of recent
studies at this laboratory3,4. We have measured the level of
calcium inclusion in barite which is precipitated from brines
containing increased [Ca2+] as shown in Fig. 11. Levels of
calcium inclusion between ~4% and ~12% are observed in our
experiments. It has been known for a long time that levels of
3
up to ~6% calcium in barite is seen for pure salts in the
absence of SI11,12. The effect of this calcium inclusion on the
size of the barite crystal a-axis is seen in Fig. 12 4 and this
confirms other published results on such measurements3. It
has also been found that3.4:
- only calcium is included in the lattice while no magnesium
inclusion is observed;
- decrease in the lattice parameter (a-axis) will probably retard
the lattice growth or at least make the lattice growth easier to
inhibit;
- even more calcium inclusion into the barite lattice is seen if
SI is present.
Previous results have also shown that SI when present is
incorporated into the barite lattice over the course of an IE
test1,7,8,9. Thus, the increased calcium inclusion and the uptake
of SI by the barite lattice are probably closely related and we
will develop this idea below.
Modelling of Ca/Mg/DETPMP System
Before presenting the full proposed mechanism for barite
inhibition by the various types of SI, we will pause to study
the main solution species involved in the process. We have
already established that it is the dissociated species Hn-xAx-
that is primarily involved in the inhibition of barite by
DETPMP, which we denote by A for simplicity. Clearly,
calcium and magnesium may bind to species A where the
degree to which binding will occur will depend on the stability
constants, KMg, KCa and KBa. For DETPMP, these quantities
are known6 to be KMg = 6.3 x 1010, KCa = 5.0 x 1010 and KBa =
1.58 x 108 and hence the level of calcium and magnesium
binding should be quite significant13. The measurement
conditions and exact definition of these constants is not
specified, but they are still informative and suitable for
modelling purposes. The solution equilibria which occur in a
Ca/Mg/DETPMP system are as follows:
⎯⎯→
K Ca
Ca + A ←⎯⎯ Ca . A
(2)
⎯⎯⎯ → Mg . A
K
Mg + A
Mg
←⎯⎯ ⎯
Although these equations are quite simplified, they still turn
out to be extremely useful and instructive. The stability
constants, KCa and KMg, are given by:
x Ca . A x Mg . A
K Ca = ; K Mg = (3)
x Ca .x A x Mg .x A
where the xi denotes the solution molar concentrations
(strictly, the activities) of species i, where i = CaA, MgA, Ca,
Mg or A. Suppose we start with initial quantities of Ca, Mg,
SI (as A) in solution, denoted x Ca0 , x Mg0 and x A0 . It is
quite straightforward to rearrange the system of equations to
obtain the following 2 implicit equations for x Ca and x Mg :
4 K.S. Sorbie and N. Laing
x Ca0 (1 + K Ca .x Ca + K Mg .x Mg )
x Ca =
(1 + K Ca .x A0 ) (4)
x Mg0 (1 + K Ca .x Ca + K Mg .x Mg )
x Mg =
(1 + K Mg .x A0 ) inhibition mechanism, we return to the results in Fig. 9. The
Eqs. 4 can be solved numerically in a spreadsheet but the
actual quantity of interest is the equilibrium value of
( )
x CaA x MgA , i.e. the ratio of SI bound to Ca to SI bound to
Mg. For very large values of stability constants KMg, KCa, as is
the case here, an approximate equation for this ratio is given
by:
⎛ x CaA ⎞ ⎛ K Ca .x Ca0 ⎞ inhibits the crystal growth. That is, it is not simply the
⎜⎜ ⎟⎟ ≈ ⎜⎜ ⎟⎟ (5)
⎝ x MgA ⎠ ⎝ K Mg .x Mg0 ⎠ crystal growth retardation. There is more than this going on in
In fact, for such high binding constants, this is an excellent
approximation. To illustrate this, consider the simple example
near the MIC; [SI] = [A] = 25 ppm ≈ 10-4 M, [Ca2+] = 800
ppm = 0.02 M, [Mg2+] = 1000 ppm = 0.042 M. This leads to
( x CaA x MgA ) ≈ 0.3779 from Eq. 5 above (exact value 0.3781
from Eq. 4) which shows that 27% of DETPMP has Ca bound
to it and 63% is bound to Mg. Another consequence of such
high binding constants is that, if there is a molar excess of
calcium and/or magnesium to SI, there is virtually no unbound
inhibitor molecules. For example, even for [SI] = 25 ppm ≈
10-4 M, [Ca2+] = 20 ppm = 0.0005 M, [Mg2+] = 20 ppm =
0.00083 M, we calculate that ( x CaA x MgA ) = 0.48, implying
that 33% of DETPMP has Ca bound to it and 67% has Mg and
the level of free SI (no Ca or Mg bound) is a negligible ~ 1.4 x
10-6 %.
We now make the assumption that only the Ca-A complex
takes part in the scale inhibition mechanism and that Mg-A is
not involved. The reason for believing that this is even
approximately true will emerge as we explain the inhibition
mechanism in more detail. In reality, the Ca-A complex is
probably much more involved than Mg-A but from the
extreme assumption above we can calculate the equivalent
activities of a given threshold concentration of SI in the
presence and absence of Mg. Using Eqs. 4 and 5, we illustrate
this from the following model example; [SI] = 4.17 x10-5 M
(ca. 20 – 25 ppm), and levels of [Ca2+] = 400, 800, 1200 and
2400 ppm, KCa and KMg are taken as the binding constants
given above. For any of these condition and for a given
[Mg2+], we can calculate the amounts of SI present (in ppm)
which is bound to the Ca, i.e. according to our assumption this
is the “active” part of the SI since Mg-A shows no inhibition
activity. Results are shown for a sequence of such
calculations in Fig. 13. This figure can be understood using
the simple example of [Ca2+] = 800 ppm and [Mg2+] = 600
ppm; from Fig. 13, we can read off that the equivalent [SI] =
10ppm meaning that, if we have these levels of Ca and Mg
and 25 ppm of SI, the equivalent IE would be seen for a 10
SPE 87470
ppm SI solution with 800 ppm Ca but no Mg. In other words,
for this case the Mg has “tied up” 15 ppm of the inhibitor and
we will refer back to this figure in developing our general
mechanism of barite scale inhibition.
Firstly, in the light of our view on the role of Ca-A in the
data from Fig. 9 is presented in Fig. 15 but is now plotted
against the [Ca]/[SI] molar ratio. This diagram holds an
important clue to the DETPMP inhibition mechanism. For
example, at the huge molar excess of calcium at [Ca]/[SI] =
600 (where [Ca2+] = 750 ppm), the IE is still only ~20%. We
know from our calculations that the inhibitor is fully saturated
with calcium. Therefore, the mechanism of inhibition for the
DETPMP is not simply that it binds to Ca which then better
presence of Ca-A (i.e. Ca – DETPMP) per se which leads to
the system and this is developed below. As a more specific
example, from Eqs. 4 and 5, we can easily show that for [SI] =
20 ppm, [Ca2+] = 20 ppm and [Mg2+] = 0 ppm, virtually
100% of the SI is present as Ca-SI complex but DETPMP
does not inhibit under these conditions. The DETPMP
requires much higher levels of Ca (factors of 100s of molar
ratios) in order to function efficiently.
Mechanism of Barite Inhibition
It is known that mineral scale formation occurs through
nucleation and subsequent crystal growth stages. Scale
inhibitors may prevent or retard each of these processes; that is
they may operate mechanistically through:
(i) a nucleation inhibition mechanism; or
(ii) a crystal growth mechanism.
Note that a third scale inhibition mechanism, dispersion, has
been proposed which helps prevent scale deposition by
inhibitors in a flowing system14. Although this is certainly a
plausible mechanism, we do not consider it further here.
Nucleation Inhibition
As noted above, all SI species take part in both of the above
mechanisms to some degree but different SIs often work
predominantly through one particular mechanism. It has been
established that polymers mainly operate as nucleation
inhibitors and phosphonates mainly work as crystal growth
retarders/blockers. This difference in inhibition mechanisms
between SI types leads to differences in performance in
dynamic (tube blocking) and static (bottle) scale inhibition
efficiency tests as discussed in Ref. 1. Denoting the MIC level
in the static and tube blocking tests1,2,4,15 as MICST and MICTB,
respectively, we can define a quantity which we call the
Performance Quotient (PQ) as, PQ = (MICST/ MICTB). To be
meaningful, the same conditions of brine composition,
temperature and pH should be used in the static and tube
blocking tests used to measure the PQ for a given inhibitor.
PQ values for different inhibitors may only be compared when
they have been measured under the same conditions. This
quantity is shown schematically in Fig. 14 where, if the PQ ~1
SPE 87470 How Scale Inhibitors Work: Mechanisms of Selected Barium Sulphate Scale inhibitors Across a Wide Temperature Range
then it implies that the species operates mainly via a crystal
growth mechanism (as for DETPMP). The higher the value of
PQ, then the more likely the species will operate through a
nucleation inhibition mechanism (e.g. for PVS). Data from
Refs. 1 and 15 have been used to calculate the PQ for various
species and results are summarised in Table 2; PQ values for
PVS, PPCA and DETPMP are 7.5, 3.6 and 1.2, respectively.
The PQ of PVS indicates that it has the largest tendency
towards operating through a nucleation inhibition mechanism.
This is consistent with our earlier finding that PVS has a high
IE at 5oC even for very high Sp brine mixtures. DETPMP has
PQ ≈ 1.2 showing much less of a tendency to nucleation
inhibition and, consistent with all of our observations, PPCA
has a PQ ≈ 3.6, which is “intermediate” between the other two
species. Therefore, our interim hypothesis is that the PQ value
gives a strong indication of the balance between the two
mechanisms for a given SI species. The indication that
DETPMP operates mainly through a crystal growth retardation
mechanism will now be developed since there are still details
of this process to establish.
Crystal Growth Retardation
We now present the various steps in our proposed mechanism
for crystal growth retardation of barite in the presence of Ca
and Mg, which is most relevant to DETPMP. We have
already indicated (Fig. 15) that it cannot simply be the Ca-A
complex which then goes on to inhibit more effectively
although we will see that this species in certainly involved in
the inhibition mechanism.
A schematic of our proposed 2-step mechanism is shown
in Figs. 16(a) and 16(b) and the discussion here will refer to
these figures as follows:
Step 1: Fig. 16(a) shows the formation of barium sulphate in
the presence of calcium and magnesium ions in the scaling
brine solution. It is known that some calcium inclusion into
the barite lattice occurs but to have a significant proportion of
Ca inclusion, the [Ca2+] must be relatively high. This is an
important central feature of our proposed mechanism; that is,
if we only have very low Ca levels (say < 100 ppm) in
solution, the lattice does not include a significant quantity of
calcium and hence it does not distort very significantly. No
magnesium is included in the barite lattice.
Step 2: Fig. 16(b) shows the situation where a threshold level
of inhibitor (say, 5 to 25 ppm) is present but with a molar
excess of Ca and Mg ions in solution (in some appropriate
Ca/Mg ratio). From our calculations, all the SI is saturated
with divalent cations with typically significant quantities of
both Ca-A and Mg-A in solution (according to Eqs. 4 and 5).
The Ca-A complex is the “active” species in terms of scale
inhibition and it is this species (only) that interacts with the
growing barite crystal and is effective incorporated with the
Ca into the lattice as shown schematically in Figs. 16(b). The
role of the magnesium is effectively to remove the active SI as
Mg-A stopping it from inclusion in the barite lattice. The
incorporated Ca-A complex both increases the amount of
5
calcium inclusion, further distorts the lattice (as seen in the
barite crystal a-axis reduction) and also explains why SI in lost
from solution as observed experimentally1,7,9. This is the key
step in the crystal growth retardation mechanism through
which DETPMP operates. This mechanism explains why the
calcium must be at a fairly high level relative to the DETPMP
(in molar excess terms) in order to work effectively. This also
clarifies the important role that the Ca-A complex plays while
explaining why it requires such high levels of calcium in
solution to be effective. The negative effect of the magnesium
on DETPMP is due to the non-participation of the Mg-A
complex in the inhibition mechanism as magnesium is not
included in barite crystal growth.
Summary and Conclusions
Summary: In this paper, we develop a more detailed view of
the mechanism of scale inhibition by drawing together
evidence from a wide range of sources, presenting a number of
new experimental results and performing some simple SI-
metal complexation calculations. This work further elaborates
on earlier views of scale inhibition mechanisms1. We
conclude that the two principal mechanisms of scale inhibition
are as previously thought, nucleation inhibition and crystal
growth retardation (while conceding that other mechanisms
may exist during flow and at surfaces14). Inhibitors working
primarily through nucleation inhibition, such as PVS (and
other polymers), show a high Performance Quotient, PQ =
(MICST/ MICTB). This explains why PVS works very well at
lower temperature (5oC) where, despite the much higher levels
of Sp, the kinetics dominate the barite formation. Here, the
degree to which the mechanism is balances between the two
mechanisms is indicated by the value of PQ. DETPMP has a
lower PQ (∼1.2) showing that it works mainly through a
crystal growth retardation mechanism and, consistent with the
experimental observations, PPCA has an intermediate value of
PQ (3.6). We have revisited this mechanism in more detail
and we have proposed a 2-step mechanism as outlined above
and detailed in the Conclusions. This mechanism explains
quite satisfactorily all of the observations and sensitivities for
the effects of pH, calcium and magnesium on inhibition
efficiency. The conclusions from this work are quite general
over a range of different scale inhibitor types. The important
issue for a given SI species is to establish what “balance of
mechanisms” it operates through. The value of the PQ, the
temperature behaviour and the sensitivities to Ca and Mg are
all indicators of this balance.
Conclusions: The specific conclusions of this work are now
presented with reference to the 3 inhibitors studied here (PVS,
DETPMP and PPCA) as follows:
(i) PVS is characterized by a high PQ and works principally
by nucleation inhibition which takes particular advantage of
the slow kinetics at low temperature (5oC) where, although Sp
is very high, inhibition efficiency is good.
6 K.S. Sorbie and N. Laing
(ii) The low pKa of PVS and the lower stability constants
between the –SO3H group and Ca and Mg explain why PVS
works at lower pH values (2) and it is not very significantly
affected by these cations.
(iii) PVS almost certainly also contributes to crystal growth
retardation to some degree but this is not its principal
mechanism of operation. The results in Fig. 8 show that some
beneficial effect of calcium is seen for PVS (and PPCA). This
may be due to the calcium inclusion in the barite lattice
causing some distortion hence making the barite rather easier
to inhibit.
(iv) DETPMP is characterized by a lower PQ value and
principally operates through a crystal growth retardation
mechanism and it is quite significantly affected by the
presence of Ca and Mg ions in solution.
(v) The higher pKa range for DETPMP (mean pKa ~4.5)
explains why it does not operate at lower pH values (< 4.5).
(vi) The higher binding constants (KCa and KMg) for DETPMP
explain some of the sensitivity to Ca and Mg seen for this
species. Calculations show that typically all of the DETPMP
(at MIC levels) is complexed to Ca and/or Mg and that there
are often significant amounts of Mg-A present if Mg is present
in solution. We postulate that only the Ca-A complex is
involved in the inhibition mechanism and any SI (A) present
as Mg-A is essentially “lost” and does not participate in the
inhibition process. This may be a rather drastic assumption
and we expect that the presence of Mg-A simply “significantly
reduces” the ability of the SI to inhibit.
(vii) The inhibition mechanism for DETPMP cannot simply be
due to the formation of the Ca-A complex leading to a better
surface docking of the SI at the barite surface i.e. it is not
simply a better lattice matching by the Ca-A relative to the A
itself. At 20 ppm Ca (0 Mg) and ~ 20 ppm of SI, there is
100% Ca-A present but it does not inhibit barite. Also, there
is no correlation with SI adsorption on barite and IE10.
(viii) A 2-step mechanism is proposed for the crystal growth
retardation mechanism of DETPMP inhibiting barite. Firstly,
a high enough calcium concentration must be present in
solution for significant Ca inclusion to occur (some 100s of
ppm) in the lattice. This causes some lattice distortion which
in turn makes it easier to inhibit. The inclusion occurs by
interaction between the Ca-A complex and its subsequent
inclusion in the barite lattice. This explains both the higher
levels of Ca inclusion when SI is present and also why the SI
in incorporated into the lattice3,4. The non-interaction of the
Mg-A complex is due to the fact that, when no SI is present,
Mg is not included in the barite lattice3,4.
(ix) PPCA shows behaviour that is “intermediate” to that of
PVS and DETPMP i.e. it shows a more even balance between
SPE 87470
nucleation inhibition and crystal growth retardation. For
example (Fig. 5), the order of effectiveness at 95oC is
DETPMP > PPCA > PVS and at 5oC this is PVS > PPCA >
DETPMP. Completely consistent with this observed
“intermediate” behaviour, the PQ of PPCA is ≈ 3.6.
(x) The effect of calcium and magnesium on the IE of PPCA –
whether positive or negative - is rather less than on DETPMP
since there are 5 -PO3H2 groups in DETPMP and just 1 -
POOH group in PPCA and the Ca-and Mg-phosphonate
stability constants are probably much higher than those
between these metals and the –POOH group.
Nomenclature
A generic “mean” dissociated scale inhibitor species,
Hn-xAx-
Ca-A calcium-scale inhibitor complex (A = DETPMP)
Mg-A magnesium-scale inhibitor complex (A = DETPMP)
DETPMP diethylene triamine penta (methylene phosphonic
acid)
HnA model n-protic scale inhibitor molecule
IE inhibition efficiency (%)
KBa barium-DETPMP stability constant
KCa calcium-DETPMP stability constant
KMg magnesium-DETPMP stability constant
MIC minimum inhibitor concentration for some defined
level of inhibition efficiency
MICST MIC in a in a static (bottle) test
MICTB MIC in a tube blocking test
PPCA poly phosphino carboxylic acid
PQ Performance Quotient, PQ = (MICST/ MICTB)
PVS poly vinyl sulphonate
SI scale inhibitor, here - DETPMP, PPCA, PVS
Sp barite supersaturation
T temperature (oC)
xi solution molar concentration (strictly, activity) of
species i; where i = CaA, MgA, Ca, Mg or A.
Acknowledgements
The sponsors of the Heriot –Watt University, Flow Assurance
and Scale Team (FAST) JIP are thanked for their support,
input and permission to publish this work: Baker Petrolite,
BioLab, BP, ChampionServo, ChevronTexaco, Clariant Oil
Services, ConocoPhillips, Exxon/Mobil, Halliburton, Kerr-
McGee, Marathon, MI Production Chemicals, Norsk Hydro,
Nalco, Petrobras, Rhodia, Saudi Aramco, Schlumberger, Shell,
Statoil, Total and Yukos.
References
1. Sorbie, K.S., Graham, G. M. and Jordan, M.M., "How Scale
Inhibitors Work and How This Affects Test Methodology",
Paper and Keynote Address: Proceedings of the 4th
International Conference and Exhibition - Chemistry in
Industry 2000, Manama, Bahrain, 30 Oct. - 1 Nov. 2000.
SPE 87470 How Scale Inhibitors Work: Mechanisms of Selected Barium Sulphate Scale inhibitors Across a Wide Temperature Range 7

2. Oilfield Scale Research Group (OSRG), Laboratory Different Barium Sulphate Oilfield Scale Inhibitors",
Procedures Manual, Heriot-Watt University, 1995. SPE81825, SPE Production and Facilities (SPEPF), pp. 28-44,
February 2003.
3. Hennessey, A. and Graham, G.M., ”The Effect of Additives
on the Co-crystallisation of Calcium with Barium Sulphate”, J. 10. Laing, N., Graham, G.M. and Dyer, S.J., “Barium Sulphate
Cryst. Growth, 237-239, pp. 2153 – 2159, 2002. Inhibition in Subsea Systems – The Impact of Cold Seabed
Temperatures on the Performance of Generically Different
4. Liang, N., “The Performance and Mechanisms of Selected Scale Inhibitor Species”, SPE 80229, Proceedings of the SPE
Barium Sulphate Scale Inhibitors Under Various Conditions of International Symposium on Oilfield Chemistry, Houston,
Brine Composition and Temperature”, PhD Thesis, Institute of Texas, 5-7 February 2003.
Petroleum Engineering, Heriot-Watt University, 2004.
11. Grahmann, Neues Jahrb. Min., 1, 1, 1920.
5. Putnis, A., An Introduction to Mineral Science, Cambridge
University Press, 1992. 12. Deer, W.A., Howie, R.A. and Zussman, J., An Introduction
to the Rock Forming Minerals, Prentice-Hall, 1992.
6. Monsanto Technical Bulletin 53 - 39 (E) ME - 3 (7/1988),
“Dequest 2040, 2050 and 2060 Series of Multifunctional 13. Oddo, J.E. and Tomson, M.B., “The Solubility and
Metal Ion Control Agents in Aqueous Solutions ”, 1986. Stoichiometry of Calcium-Diethylene triamine penta
(methylene phosphonate) at 70oC in 2.0M Brine Solutions at
7. Graham, G.M., Boak, L.S. and Sorbie, K.S., "The Influence 4.7 to 9.0 pH”, SPE 20378, 1990.
of Formation Calcium on the Effectiveness of Generically
Different Barium Sulphate Oilfield Scale Inhibitors", 14. Yuan, M.D., Jamieson, E. and Hammonds, P.,
SPE37273, Proceedings of the SPE International Symposium “Investigation of Scaling and Inhibition Mechanisms and the
on Oilfield Chemistry, Houston, Texas, 18-21 February 1997. Influencing Factors Iin Static and Dynamic Inhibition Tests”,
Paper 98067 prepared for presentation at the NACE Annual
8. Boak, L., Graham, G.M. and Sorbie, K.S., “The Influence Conference and Exposition CORROSION98 held in San
of Divalent Cations on the Performance of Barium Sulphate Diego, CA. March 22-27, 1998
Scale Inhibitor Species”, SPE50771, Proceedings of the SPE
Symposium on Oilfield Chemistry, Houston, Texas, 16–19 15. C. E. Anderson, “Investigation of the Thermal Stability of
February 1999. Amine Methyl Phosphonate Scale Inhibitors and Identification
of their Degradation Products”, MPhil Thesis, Heriot-Watt
9. Graham, G.M., Boak, L.S. and Sorbie, K.S., "The Influence University, 2002.
of Formation Calcium on the Effectiveness of Generically
Table 1: Brine compositions
Brine Compositions in mgL-1 used in Static Efficiency Tests
Ion Synthetic Brent FW Forties FW
Seawater
Sodium 10887 6350 29370
Calcium 428 145 2809
Potassium 1368 25 504
Magnesium 460 140 372
Barium 0 20 252
Strontium 0 25 574
Sulphate 2960 0 0
Chloride 19768 10279 52360
TDS 35871 16984 86241
Ionic Strength / M 0.73 0.30 1.65

Table 2: Table of Peformance Quotients (PQ = MICST/MICTB) for various scale inhibitors
Tube blocking 2 hour static Performance
MIC / ppm MIC / ppm quotient
(MICst/MICtb)
PVS 1 7.5 7.5
Hex-P 0.75 3 4.0
PPCA 2.75 10 3.6
Ter-polymer 2.5 8.5 3.4
DETPMP 6.5 8 1.2
8 K.S. Sorbie and N. Laing SPE 87470

PO3H2
P 3H2
PO
OHH
Calculated speciation of DETPMP as function of pH
O OH O O
N O H(n-x)Ax-
N N PO3H2 H10A x = 3, 4 mainly at pH 4 A10-
8
P 9

H H 3
PO3H2 PO3H2 m n
OH 1
2
4
7
5
6
Diethylenetriamine penta (methylphosphonic acid) (PPCA))
Phosphino polycarboxylic acid (PPCA
- DETPMP

H
H
H n
O
S
HO
O

Polyvinyl sulphonate - PVS

Fig. 2: Speciation of DETPMP vs. pH

Fig. 1: Structures of scale inhibitors used in this work

Barium sulphate inhibition efficiency of inhibitors at 5C, 50C
and 95C after 22 hours in 50:50 Brent FW:SW
100
DETPMP PPCA PVS
(a) BaSO4 inhibition efficiency– DETPMP, 80 5C
PPCA and PVS vs. time; pH 2, 95oC 50C
% Efficiency

95C
60

40

20

0
0.5 1 3 0.1 0.5 1 0.1 0.5 1
Inhibitor conc / ppm

Fig. 4: BaSO4 inhibition efficiency of DETPMP, PVS and

PPCA vs. Temp.; 50:50 Brent/SW brine mix after 22hours.

Barium sulphate inhibition efficiency of inhibitors at

5C, 50C and 95C after 22 hours in 50:50 Forties FW:SW
100
(b) BaSO4inhibition efficiency – DETPMP, DETPMP PPCA 5C PVS
PPCA and PVS vs. time; pH 7, 95oC 50C
80 95C
% Efficiency

60

40

20

0
10 20 30 10 20 30 10 20 30
Inhibitor conc / ppm

Fig. 5: BaSO4 inhibition efficiency of DETPMP, PVS and

PPCA vs. Temp.; 50:50 Forties/SW brine mix after 22 hours.

Fig. 3: BaSO4 inhibition efficiency at pH 2

and pH 7 for DETPMP, PVS and PPCA [Ref. 1]
 SPE 87470 How Scale Inhibitors Work: Mechanisms of Selected Barium Sulphate Scale itoAcross a Wide Temperature Rang 9

Inhibition efficiencies after 22 hrs. Forties type brine Inhibition efficiencies after 22 hrs. Forties type brine
Iso-supersaturaton = 310 Iso-supersaturaton = 3189

Fig. 6: Isosupersaturation IEs (Sp = 310); Forties/SW brine mix

Fig. 7: Isosupersaturation IEs (Sp = 3189);Forties/SW brine mix

Barium sulphate inhibition efficiency vs. [Ca2+] at 95oC, pH 5.2,

2 ppm active SI, 2 hour residence time
Figure 4.241 Effect of Increasing [Ca] on Inhibition Efficiency
of DETPMP and PVS at 95C
100

PVS
80
(2 & 22 hours)
% Efficiency

60

40
DETPMP DETPMP 2h
(2 & 22 hours) PVS 2h
20 DETPMP 22h
PVS 22h
0
0 200 400 600 800 1000 1200 1400 1600 1800
[Ca] / ppm

Fig. 8: Barium sulphate IEs vs. [Ca2+], 95oC [Refs. 1, 7,9]. Fig. 9: IE (%) for DETPMP and PVS vs. [Ca2+], 95oC.

BaSO4 Efficiency (%) vs. [Mg2+]; [SI] = 8ppm, 95oC; pH 5.4; Calcium inclusion (Ca/Ba) Molar ratio vs. initial [Ca2+]
60 SSW:40 MSFW; Residence time 2 hours
0.14

Medium scaling Forties

0.12
type water
0.10
BaSO4
Efficiency (Ca/Ba) 0.08
5oC
(%) molar ratio 50oC
0.06

0.04 90oC

0.02
Initial [Ca2+]
0.00
Magnesium conc. (ppm) Æ

Fig. 10: IE (%) for PPCA, DETPMP and PVS vs.

Fig. 11: Calcium inclusion, Ca/Ba molar ratio vs. initial [Ca2+].
[Mg2+], 95oC.
10 K.S. Sorbie and N. Lain SPE 87470

a-axis ( Å) as function of calcium inclusion (Ca/ Ba Molar ratio) Prediction of equivalent [SI] with zero Mg for
same Efficiency
30

Equivalent effect [SI] wit no

25
[Ca2+] (ppm)
20

Mg
15 2400

10
1200

5 800
400
0
0 500 1000 1500
[Mg] ppm

Fig. 12: a-axis (Å) as function of calcium inclusion measured Fig. 13: Model prediction of “available” [SI] vs. [Mg2+] at
as Ca/Ba molar ratio, [Ref 4]. various values of [Ca2+] (see text).

Performance Figure
Effect 4.242 Effect
of [Ca]/Inhibitor of on
ratio Increasing
inhibition Initial [Ca]/[inhibitor]
efficiency of DETPMPRatio on at 95oC
and PVS
MICST
quotient = Inhibition Efficiency of DETPMP and PVS at 95C
MICTB 100
PVS
(2 & 22 hours)
MICST = 1 80
MICTB
MICTB
% Efficiency

60
Tube
blocking DETPMP
40
DETPMP 2h
DETPMP
(2 & 22 hours) PVS 2h
20
DETPMP 22h
PVS PVS 22h
0
0 200 400 600 800 1000 1200 1400
MICST 2 hour static
[Ca] / [Inhibitor] (molar ratio)

Fig. 14: Schematic illustrating the physical meaning of Fig. 15: IE (%) for DETPMP and PVS vs. [Ca2+]/[SI] molar
Performance Quotient (PQ) . ratio, 95oC; same data as in Fig. 9.

SALTS Ca2+ & Mg2+

+ SI in solution
Ba2+ Ca-SI
Ca substituted incorporated
SO42- SALTS ONLY
Ca2+ & Mg2+ in BaSO4- up to
Ca2+ solution ~6% Ca - no Mg
Mg2+
SI

Normal Normal
BaSO4 BaSO4 Ca/SI substituted BaSO4
- more Ca bound by complexation

Fig. 16: Schematic of the proposed barite crystal growth mechanism for DETPMP showing (a) No SI: normal BaSO4
growth with Ca inclusion and no Mg inclusion; (b) SI present: BaSO4 growth with Ca-A complex inclusion and Ca
inclusion and no Mg-A inclusion (all SI complexed as Ca-A and Mg-A); A = SI, in appropriate dissociation state (see text).