Separation in Oilfield

Operations
Gunbarrels: Myths versus Realities

A Technical Paper
Prepared for

All Interested Oilfield Personnel

A Paper by:

High-Tech Consultants, Inc.

HTCs Cold Weather
HWSB Skim Tank
Patented 2011

6840 East 112th Street South

Bixby, OK 74008-2062
Office Phone: 918-298-6841
Cell Phone: 918-231-9698
Author:
Bill Ball, President
Email: [email protected]
August 1, 2013

High-Tech Technologies

Proven Oilfield Separation Systems

are more efficient than others.

paper discusses a few of them.

PREFACE
Knowledge is everything! With oil prices
hovering around $100/barrel, its not
surprising that the oil industry has a
penchant for producing and selling
every drop. Yet, there is a very large
knowledge gap regarding how to
make this happen. This paper attempts
to fill some of that gap, improving the
readers understanding of how to
capture every drop of oil in todays
operations. It attacks the myths of the
past, myths that tend to perpetuate
themselves in the we always did it that
way mindset we encounter every day.
It prompts us to become better at what
we do so we can meet our goals of
maximized
oil
revenue,
facilities
performance, and profitability.

BACKGROUND HISTORY
The practice of oil-water separation
began with the first oil well in Titusville,
PA over 150 years ago out of necessity.
Its the same today.
Crude oil
producers must separate produced gas,
solids, and water from oil in order to sell
it and avoid purchase price penalties.
As an industry, we also need to
separate contaminants from produced
water to minimize disposal or injection
well plugging.
These separation
processes can be accomplished with
many different types of equipment and
in many different ways. Some of these

This

In any separation system, the first

needs are to select the proper size and
the right design for the appropriate
separation process.
Before deciding
what the proper size and the right
design are, it is useful to know how the
more common and widely accepted
designs evolved.
For the last 15 decades or so the oil
producing industry has used a design
which is known as the atmospheric
gunbarrel, or wash tank, to separate
and sell crude oil. In this design the
focus was on removing water and BS&W
from produced crude oil, with no
concern for water quality or oil
carryover whatsoever.
In this system produced oil, water, gas
and any solids first flows into a
degassing chamber referred to as the
gas boot.
Gas flows up and is
equalized with the gas phase in the tank
through a separate pipe. It commingles
with gas evolving from the crude and is
piped off to a flare, vapor recovery
systems and on to sales, or in many
cases to an atmospheric vent.
Liquids flow down and out of the gas
boot into a downcomer pipe which
ends near the tank bottom where the
liquids exit under a baffle, or spreader.
2013 HTC, Inc.

This document was prepared by HTC

for the customers and clients of KBK Industries and HTC, Inc.

Separation In Oilfield Production Operations

August 1, 2013
Page 3

Myths versus Realities

Many spreaders were designed with

serrated bottom edges. These were
inverted V shaped notches, like saw
teeth. The v-notches were intended to
meter or distribute the flow of oil
uniformly through the water phase. The
design philosophy was that washing
the crude oil through water would allow
the water to flow out of the crude oil.
Thus the name wash tank.
Some designs placed the gas boot and
downcomer outside the tank, while
others placed the gas boot on top of the
tank (see Figure 1 on Page 3). In both
cases, when viewed from the top, the
tank and gas boot circles appeared
similar to the circles one sees looking
down the barrel of a gun. This was the
basis for the term gunbarrel. The
gunbarrel, or wash tank, design was
originated in the 1860s. Trial and error
applications saw its size increase larger
and larger until it finally performed as
desired. This propagated the oilfield
design philosophy that bigger is
better.
For better or worse, this simple design
remained unchallenged for decades. It
was and is a standard of the industry for
atmospheric crude oil dehydration. But,
as well see, it is unfortunately very
inefficient!
The Gunbarrel design was finally
challenged in the 1970s, when the
uncommonly high price of crude oil
justified field and laboratory research
into the relative efficiency of this and
other standard oilfield equipment

designs. The results were startling, and

most were disappointing!
Retention time studies were conducted
in separators, gunbarrels, wash tanks,
free water knockouts, heater treaters,
and storage and skim tanks. The results
showed that the hydraulic efficiency of
all designs were extremely low, most
ranging from 1% to 3%, with a few
ranging to only 20%. Bigger had been
thought to be better, but these studies
refuted this concept. In most cases,
bigger was proven to be worse!

GAS

OIL

WATER

TYPICAL
GUNBARREL/WASHTANK
FIGURE 1

2013 HTC, Inc.

This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 4

Myths versus Realities

These studies also identified the flow

characteristics
leading
to
low
efficiencies, and helped explain some of
the
reasons
for
these
gross
inefficiencies.
These studies also proved that all fluids
predictably take the path of least
resistance. They traverse in a short,
narrow flow path between the vessel
inlet and outlet. This was something of
a revelation since it was presumed that
fluids naturally distributed uniformly in
these vessels, and flowed like a plug or
piston through the entire cross section.
In addition, these studies proved that
flowing path and velocity determine the
degree or efficiency of separation, and
that they are almost never constant in
the real world of oilfield operations.
What was constant was that when the
velocity of the bulk production exceeds
the separation velocity of any fluid or
solid, the mixing energy offsets or
eliminates most or all separation.
This was a revelation. It confused many
of those involved in these studies. After
all, oil industry separation designs had
always been assumed to work
efficiently!
As this new level of knowledge was
applied
to
traditional
equipment
designs, the gunbarrel (wash tank)
results were also surprising. In the
past, gunbarrel sizing had been based
on the supposition that oil needs eight
to 24 hours of piston displacement
retention time to dehydrate, depending
on API gravity. Decades of observation
had determined that these retention

times could be related to the gravity of

crude, from 20 API to 40API. It was
known that in heavier oil applications,
all bets were off, as much more time
was needed. This had meant larger and
larger tanks.
Those same studies showed that the
metering concept of the serrated 9sawtooth) distributor baffles was invalid too.
In fact, it was found that crude oil
entering a gunbarrel (and most other
vessels) oil wets ALL of the surfaces,
including the serrated spreader, and
then, rather than being metered out into
the water phase by the serrations, the
crude was found to really flow in rivulets
attached to the steel spreader structure,

GAS

OIL

WATER

CRUDE FLOW PATH

FIGURE 2

2013 HTC, Inc.

This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 5

Myths versus Realities

up the outside of the spreader to its

crest where if travelled up the OD skin
of the downcomer and into the oil layer
above, or disengaged from the edge of
the spreader and into water layer. If
the tank was designed like the one in
Figure 2, oil disengaged at the highest
or sharpest edge surface in large
droplets or globules. These oil droplets
then flowed rapidly and vertically into
the oil phase. From there very little, if
any, horizontal distribution occurred,
and the oil flowed diagonally to the
outlet nozzle, circumventing the bulk of
the tanks volume. It was found that
only a very small portion of the storage
volume actually came in contact with
the crude. In many cases the inlet
crude was found to actually reside in the
oil phase for just a few minutes. This
finding
refuted
the
old
design
philosophy of sizing based the 8-24
hours of crude oil retention time rule of
thumb.
It became clear that any water in the oil
phase had to separate in a matter of
minutes, rather than in a matter of
hours. And, since the age-old belief
that 8-24 hours of retention time was
proven incorrect, it became obvious that
gun barrels had to have been vastly
oversized to work, given the extremely
low hydraulic efficiency of their design.
At this point, design emphasis began to
shift away from the assumed oil
retention time of the past.
Some
designers began to look for new
methods of increasing the actual
distribution of oil within a smaller
vertical column (volume) of stored crude

to improve dehydration. It was believed

that this would result in smaller
gunbarrel designs, and a smaller
investment in the oil stored in them.
Lucite models were constructed so small
scale lab tests of various internals could
be performed. In general terms, the
results were quite encouraging.
And not surprisingly, each investigator
was amazed at how inefficient the old
original gunbarrel design actually is (see
Figure 3).
It became clear that when fluid flows
and distribution were more uniform
distributed the results were a dramatic
increase in actual crude retention time.
As these studies progressed it was
eventually discovered that truly uniform
distribution of incoming crude and

OIL

Gas
Oil

WATER
ORIGINAL GUNBARREL
DESIGN
Figure 3

2013 HTC, Inc.

This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

emulsion through the entire cross

section of the oil layer increased
retention times by a factor of 10 to 35.
Furthermore, and equally surprising, it
was also found that process capacity
and separation efficiency increased even
more when the inlet crude and emulsion
is NOT washed through the water
phase, but is introduced into the
emulsion rich layer just above the water
phase. These studies proved that both
naturally occurring and artificially added
emulsion breaking chemicals tend to
concentrate in this layer.
These
chemicals have the effect of reducing
the surface tension of droplets,
promoting coalescing (the growth of
droplet size), and therefore vastly
improving separation.
Finally, it was observed that flow
velocities needed to be carefully
considered. A velocity relationship was
observed wherein a fractional oil flow
rate directly proportional to API gravity
accomplished the desired gravity
separation of water from oil.
This
brought a whole new engineering
dimension to the concept of proper
design.
The conclusion of this work proved that
new wash tanks/gunbarrels can be sized
much smaller, reducing the capital
investment for tankage. The investment
of
oil-in-inventory
in
smaller
gunbarrels is also obviously reduced.
Further, these high-tech gunbarrels
produce a very high quality oil effluent
regardless of summer-winter swings in
temperature and viscosity, for most oils

August 1, 2013
Page 6

in the 27 API and above range.

However, since the separated water
short circuits to the outlet, the quality of
effluent water in these gunbarrels was
shown to still be extremely poor. It
typically ranged from 1500 ppmv to
5000 ppmv oil in the effluent water, or
more.
Water quality had been ignored in the
early decades of the industry, but with
the advent of water flooding in the late
1940s, it took on a new meaning. When
produced water contained so much
remnant oil it was found to be plugging
injection and disposal wells, and as the
price of crude rose from its government
regulated pre-war $1.25/barrel level to
an unregulated and much higher value,
the issue of removing oil from water,
and the issue of water quality in
general, finally reached a level of
significance to producers.
As the industry entered the boom
years of the 1970s, several new
equipment designs were developed to
improve water quality. These designs
experimented
with
matrix
plate
coalescing, deep bed sand-based
coalescence, Lamella plate separators,
serpentine vane coalescing, and a wide
variety of other internals and processes.
All of these showed promise, but when
applied in the real world, often proved
to be impractical due to high capital
costs, premature plugging of internals,
and other related costs.
Then came the bust years of the
1980s when the concept of survival in a
depressed industry outranked the need
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 7

Myths versus Realities

for research. As oil prices sunk to

record lows many of the best minds in
the oil industry left, seeking work in
other, less volatile industries. And with
them departed a good deal of the
industrys knowledge, never to return.
In contrast, HTC persevered in its effort
to continue research and to develop
better, more efficient systems. In 1993
these efforts resulted in a series of
patents finally advancing the design
engineering of gunbarrels, wash tanks,
skim tanks, de-sanders, flow splitters,
and many others into the realm of the
highly efficient. These designs were
honed throughout the 1990s to become
performance standards in the new
millennium evolving finally into a
family of proven and widely accepted
21st century HTC design technologies.
SEPARATION AND STOKES LAW
The static separation of immiscible fluids
(fluids that are not soluble in one
another), and/or suspended solids, can
be predicted by applying Stokes Law of
physical separation. Predicting static
separation is very straight forward. An
example is predicting the separation of
gravel dumped into a tank of water.
The tank is static, which means there
is no motion inside. By applying Stokes
Law anyone can calculate how long it
will take for the gavel to reach the
bottom of the tank. It is obvious that
the gravel will settle to the bottom
because gravel is heavier than water. It
is logical that the larger, heavier pieces
of gravel will settle (separate) faster,
and the smaller, lighter pieces will settle

(separate) slower. An understanding

this simple principle is a good beginning
to understanding gravity separation
and Stokes Law.
However, the word static is the key to
distinguishing the merits of Stokes Law
from the dynamic separation typically
demanded in oilfield separation systems
where fluid stays in motion all the time.
Most oilfield process separation systems
are not static. There is constant if
irregular motion inside process vessels.
Nothing is static! So, a law that predicts
the rate of static separation had to be
modified for oilfield operations where
almost nothing is static.
Fluids are
constantly flowing all the time. Because
of this, Stokes Law does not go far
enough by itself to be applied to most
process separation challenges in our
industry.
Lets review. A typical oilfield process
separation system can be accurately
described as those with continuously
flowing conditions where all fluids are in
motion. But, Stokes Law only predicts
separation in a static, non-moving
environment.
Nevertheless, a good
understanding of the concepts set forth
in Stokes Law is considered critical to
the understanding of separation. So,
we start with it.
STOKES LAW
Stokes Law was published in 1851. It
represents the velocity of a rising or
falling fluid or particle under static
conditions with the following formula:
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

August 1, 2013
Page 8

density difference of the two phases (d1

d2) times the square of the size of the
fluid/solid particle (r2) times the
gravitational constant (C), divided by
the viscosity (N) of the continuous
phase.

Where:

F = 6rv,
and where
F is the frictional force
r is the particle radius

is the fluid viscosity, and

v is the particles speed
Vs is the particles settling velocity,
g is the acceleration of gravity,
p is the density of the particles, and
f is the density of the fluid
Stokes postulated that if the particles
fall through a viscous static fluid by their
own weight, then he could derive their
settling velocity by equating the
frictional force with gravitational force.
In order to relate Stokes Law to the
dynamic
separation
problems
encountered in typical oilfield separation
it needed to be modified. A good deal
of work was necessary to accomplish
this. The modified Stokes Law can be
represented in formula form as follows:
V = Cr2(d1 d2)
N1
The Stokes Law formula focuses on two
immiscible phases at one time. When
more than two are present, each is
calculated independent of the others.
Modified Stokes Law states that the
velocity (V) of separation is equal to the

In both versions, all of the variables

have a decided impact on separation.
However, the greatest impact is the size
of the particles to be separated, since
the relationship is not linear (one-toone), but instead is exponential (the
square) of the droplet or particle size.
That is, as the particle size doubles, the
separation velocity is increased by four
times. As size triples, the separation
velocity is nine times as fast. As size
quadruples, the particle separates
sixteen times as fast, and so on. The
inverse it also true. A reduction in size
of half results in a separation time four
times longer. A reduction in size to onefourth the original size results in a
separation time sixteen times longer,
and so on.
It is very important to grasp this
concept in the real world, since many of
the ways we handle and treat produced
fluids may reduce or enlarge the size of
the particles and droplets we are
eventually going to try to separate.
Separation in the oil patch involves the
separation of oil, gas, water, and solids
from one another. Separating oil from
water is necessary to achieve the oil
quality necessary so the oil can be sold
and shipped to refineries without BS&W
penalties.
Separating oil and other
contaminants from water is necessary
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

so the water can be re-injected or

disposed of without the plugging effects
of entrained oils and solids on injection
or disposal wells. Furthermore, whereas
oil has a commercial value, any oil left in
the water not only causes injection or
disposal well plugging, but also reduces
the income stream from the sale of the
oil now lost to the water phase.
So clearly, larger particles sizes are
preferable. One of the ways we grow or
reduce the size of the droplets or solids
we are going to try to separate is with
the addition of oilfield chemicals. In
most non-chemically stabilized mixtures
of oil in water, the majority of the
droplets of one in the other are larger
than 150 microns.
These droplets
separate rapidly simply because of their
size. However, when the droplets or
particles are smaller than 150 microns,
separating them becomes a real
challenge. Over treatment with oilfield
chemicals is a major culprit reducing
sizes, and thereby in causing chemically
stable emulsions and poor separation.
Many other circumstances also cause
smaller droplets. Pumping is another
real world culprit. For instance, when
mixtures with large droplets are moved
from one place to another, the act of
moving them may provide the necessary
physical force necessary to divide the
large droplets into smaller and smaller
droplets. Imagine the impeller of a
typical centrifugal pump, turning at
3550 RPM through a mixture of
produced water and oil. The rapidly
turning impeller shears larger droplets
into smaller and smaller droplets as it

August 1, 2013
Page 9

turns, pumping the mixture down the

line.
The smaller droplets separate
slower, consistent with Stokes Law.
In most oilfield operations there are
many
circumstances
that
make
separation more difficult than it could
be. For instance, oil flows from the
subterranean reservoir at an everincreasing velocity into the well bore.
There it A) flows to the surface via its
own energy, encountering a choke
where large droplets are broken into
tiny droplets by the forces of pressure
reduction (just as in the homogenization
of milk), or B) is picked up by a plunger
or centrifugal pump which exhibits
enormous shearing forces in the process
of moving fluid through dozens of
impellers, or past the ball and seat
discharge check valve.
In these and most other cases the result
is the same droplet size reduction. Any
reduction of the size of the droplet or
particle slows down the separation
process. When the droplets/particles
are smaller than 150 microns in the real
world, the mixture is redefined as a
suspension, and separation occurs at a
snails pace.
To get a grasp on the sizes we are
discussing here, it may help to know
that one micron is 1/1000th of a
millimeter, or 1/24,400ths of an inch.
While these are very tiny particles or
droplets, a person with normal vision
can see a 75 micron particle with the
naked eye. So, while the 150 micron
threshold is tiny, it is not beyond the
boundaries of unaided human vision.
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

Before we go further, we need to

understand a few more terms.
In the world of liquid or gas separation
models, one or more contaminants are
usually suspended in a fluid (liquid or
gas) that makes up the largest
percentage
of
the
mixture
or
suspension. This larger, or majority
fluid is called the continuous phase.
Additionally, when a mixture is made up
of a continuous phase with larger than
150 micron droplets or solid particles, it
is considered an unstable mixture or
emulsion. When the contaminants are
smaller than 150 microns, the mixture is
considered a suspension. The smaller
the micron size, the more stable the
suspension, until finally the suspension
is so stable that no Stokes Law
separation occurs.
Lets focus on suspensions that do
separate. These suspensions are nonstable mixtures of the continuous and
non-continuous phase. The degree of
stability of the mixture depends mostly
on the size of the non-continuous
phase. Again, when average size of the
non-continuous phase particles is larger
than 150 microns separation will occur
readily, usually within minutes.
When the average non-continuous
droplets range from 50 to 150 microns,
separation times often increase from
several hours to several days. When
the droplets range from 10 to 50
microns, the mixture is considered
dispersion where separation may take
many days or even weeks. And, when
the particle size of the dispersed fluid is

August 1, 2013
Page 10

smaller than 10 microns, it is considered

a colloidal suspension, where separation
may not occur in any practical period of
time.
Homogenized milk is a good example of
a colloidal suspension. Milk is a mixture
of butter fats (organic oils) and water.
In the homogenization process, the
mixture of raw milk (butterfat and
water) is pumped through a tiny orifice
under very high pressure. This shears
the butter fat particles until they are
smaller than 1 micron. The result is a
stable, non-separating dispersion of two
immiscible
fluids
(please
see:
http://www.foodsci.uoguelph.ca/dairyed
u/homogenization.html
for
more
information).
This is the kind of suspension often
caused by chokes in oil/gas wells, by the
shearing action of multistage downhole
centrifugal pumps, by cavitating surface
transfer pumps, and by leaking balls and
seats in rod pumped wells.
When we look at the actual conditions
typical of most oilfield operations, we
find that most oil in water and water in
oil have particle (droplet) sizes above
150 microns when the produced fluid
reaches the surface.
These are
mixtures which should separate rapidly.
However, this is not always the case.
When a droplet is sheared from a
mixture size of 200 microns to a
suspension size of 50 microns, we know
that the rapid separation we might have
otherwise expected will not occur. If
the 200-micron droplet separated in 5
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 11

Myths versus Realities

minutes, the 50-micron droplet (now

one fourth of the original size) will take
a calculated sixteen times longer, or 625
minutes (the square of the square of the
original separation time, according to
Stokes Law). This dramatic difference
is the reason we will want to
concentrate on the methods of fluid
handling in production operations. This
helps explain why it is sometimes
difficult to separate very light oil from
very heavy produced water.
Again, poor fluid handling techniques
can cause droplet/particle shearing,
lengthening the required times for
separation.
When this happens,
producers tend to spend too much
money on oversized surface facilities.
This is usually a waste of money.
So, it pays to understand fluid handling.
A few key examples of fluid handling
mechanics that cause droplet shearing,
longer than expected separation times,
and larger (more costly) than necessary
surface facilities, are:
The flow of produced fluids through
small restrictions, like:

the ball and seat of a rod

pump, through a choke
the flapper of a check valve
a pinched flow control valve
centrifugal sub-surface and
surface pumps
surface gear pumps
trim sets in liquid level
control valves on separators,
free water knockouts, heater
treaters, etc.

Pumping produced fluids from one

vessel to another, as in:

circulating tank bottoms

drawing interfaces off of
gunbarrels
recirculating interfaces
recycling sump liquids back
to the separation processes

In addition, most chemical additives

used in oilfield operations also have the
effect of reducing particle sizes.
Examples are:
Emulsion breakers when high
instantaneous dosages are applied,
such as:

Slugging a gunbarrel to
break a difficult emulsion
Slugging a heater treater to
clean up the oil pad
Overtreating
the
entire
production steam
Overtreating a single well
steam

Corrosion
Inhibitors:
These
chemicals often depend on water
wetting surface active agents to
clean organic deposits from the
corrosion sites.
These powerful
surface active agents (surfactants)
promote very stable oil-water and
oil-water-solids emulsions.
Scale Inhibitors:
Both organic
inhibition polymers and inorganic
scale inhibitors are formulated to
disperse
solids,
preventing
agglomeration. This is the exact
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

opposite from coalescence (droplet

or particle size growth).
While
stable dispersions are not defined as
emulsions, the results are much the
same, since the dispersants prevent
coalescence (droplet or particle size
growth).
Acids:
Acids are used for well
stimulation.
By definition, acids
have very low pH values. A low pH
environment promotes dispersion.
Therefore, droplet and particle
coalescence will not normally occur
in low pH environments.
Acids
applied
in
oilfield
production
operations nearly always contain
surface-active chemicals used to
remove the oily deposits from the
reservoir rock and scale the acids
are designed to attack.
These
surfactants
promote
chemically
stable emulsions, and this problem
is enhanced further by the presence
of the very small (usually less than
one micron) solids particles carried
back to surface treating facilities by
spent acids.
Frac Polymers: Like scale inhibition
polymers, frac fluid polymers create
very difficult and stable inverse
emulsions.
These are often so
stable that they can only be broken
by breaking down the polymer
chains. This both costly and labor
intensive, since all frac companies
continuously alter their frac polymer
chemistry.
This situation is
compounded by the huge initial
amounts of polymer containing

August 1, 2013
Page 12

flowback water the producer must

scale up to treat.
Chemically stabilized emulsions add time
to the physical separation, as has been
described in the preceding explanation
of Stokes Law. This report can shed
light on the causes, but only real-world
experience can help predict increased
separation time at the Flanagan Lease.
OIL-WATER
SEPARATION
RETENTION TIME

AND

It can be said that effective physical

separation is a function of efficient fluid
distribution and time.
The required
separation time is often referred to as
retention time, or the amount of time
a fluid is allowed to reside in a process
vessel before for the desired separation
takes place.
A key factor contributing to oil-water
separation in facility design is the
prediction and determination of real
retention time.
From the above it is obvious that droplet
or particle size is the most critical factor
when attempting to predict what
retention time may be needed, since it
is the only exponential function in
Stokes Law. It is also obvious that the
required retention time must be
provided or separation will not occur.
If the flow through surface facilities
short-circuits,
separation
will
not
adequately occur regardless of size. If
the surface facilities are too small,
separation will probably not occur since
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

August 1, 2013
Page 13

sufficient time will not be available, even

if distribution is excellent. If facilities
are properly sized but the attention to
distribution is lacking, separation
efficiency will suffer.
And, if the
facilities are too large, money is wasted.

designs. These will save money upfront and during the entire life of each
lease or field.

It is clear that too much money has

been wasted on poorly designed,
oversized surface facilities throughout
the history of the oil industry.
Unfortunately, this trend has not
slowed. In fact, in boom times like
these, just the reverse is true. This
happens because the industry is limited
in human resources, so it focuses on the
larger issues, letting issues like this go
unattended.
This also has happens
because of the widespread lack of
information, and a general lack of
knowledge. The most common oilfield
approach to purchasing surface facilities
has been to simply oversize everything;
to throw money at the problem.

It is a common belief that if we produce

2000 barrels a day, and we believe we
need one-half a days worth (12hours)
of retention time to accomplish the
desired separation, we set a 1000-barrel
capacity process facility. This makes
sense, or does it?

When faced with the prospect of a new

prolific oil well, shut in and waiting for
surface process facilities, efficiency
often takes a back seat. Even today,
most surface facility designers copy
what was done the last time, particularly
if it seemed to work. This perpetuates
the mistakes of the past. So, it is
important that we understand it is
possible to find the right balance
between separation needs, retention
time, surface facilities design, and cost.
With a basic knowledge of separation, it
is possible to leverage the available
technologies and good operating
practices to optimize surface facility

RETENTION TIME THE FACTS

AND THE FICTION

Well, each year hundreds of retention

time studies are performed worldwide.
They confirm that the fact that the
actual retention time of most facilities is
only a small fraction of the design goal.
We can define the optimum design as
having 100% hydraulic efficiency.
That is, the fluid entering a facility
designed for 12 hours of retention time
leaves 12 hours after it enters.
In reality, the design goal of 100%
hydraulic efficiency is rarely approached.
When 100% hydraulic efficiency is
achieved, flow velocities are so rapid
that
mixing
occurs,
instead
of
separation. So, a hydraulic efficiency
that is too high can cause mixing rather
than separation.
Hundreds of actual field tests prove that
actual retention time in existing process
facilities, even those with the most wellknown, best liked designs, are in the
0.1-21% range of the ideal design goal
of 100% hydraulic efficiency.

2013 HTC, Inc.

This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 14

Myths versus Realities

The fact that the difference between the

design and the actual hydraulic
efficiency is so great is both
enlightening and discouraging.
An
efficiency
of
1-21%
is
totally
unacceptable for most of us, no matter
what the subject.
A better system is clearly needed, and
warranted!
In order to increase the hydraulic
efficiency in oil-water separation process
vessels the designers include special
flow distribution systems. These include
baffles, velocity increasing orifices,
torturous matrices, vortex creating
devices, parallel plates, random masstransfer materials, coalescing tubes, and
woven synthetic cloth barriers, just to
mention a few.
These and many others have a positive
effect on retention time and hydraulic
efficiency.
Most increase the actual
retention time by a factor of two to
three times, often increasing retention
time to from 1% to 3%, 3% to 6%, or
even 6% to 21%.
Some attempts have had a negative
effect on separation, too! If a design
accelerates the fluid flow in a vessel so
it is flowing faster than the Stokes Law
rise/fall rate of the separating droplets,
mixing occurs. Mixing is the opposite of
separation.
When this happens the fluids are not
sufficiently exposed to the necessary
dynamic flow conditions for separation
to occur, even though the retention time

and hydraulic efficiency may be

theoretically or realistically improved.
From this you can see that there is more
to enhancing separation than simply
increasing retention time.
Lets consider a sample water clarifier
process vessel.
When the flowing
velocity of any oil droplet exceeds
separation velocity for that droplet, all
droplets of that size or smaller fail to
separate. These droplets flow out with
the water.
The point to be made here is that
separation is interdependent of both 1)
retention time, and 2) proper fluid flow
characteristics. Unless both are correct,
separation suffers.
This is further explained by an example.
When an oil droplet flow velocity
exceeds the Stokes Law oil-water
separation velocity for that droplet, that
oil droplet (and all the same size or
smaller) is carried with the water phase
to the water outlet. The point to be
made here is that separation is
interdependent upon both 1) retention
time, and 2) proper fluid flow
characteristics. Unless both are correct
separation suffers.
AN EXAMPLE CASE
To bring all of this into focus, lets look
at an example case.
Assume the example gunbarrel is a
standard 1500 barrel of the conventional
design shown in Figure 1.
This
gunbarrel contains 1/3rd water and
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 15

Myths versus Realities

2/3rds oil. Oil production is now 20

barrels per day. Therefore, the 20 B/D
oil flows through 1000 barrels of stored
oil in the gunbarrel. By dividing the oil
capacity of the 1000 barrel gunbarrel by
the 20 B/D production, it is easy to see
that the ideal oil retention time is 50
hours.
In our example the ideal crude oil
retention time is 50 hours based on the
tank volume versus flow per day
formula.
To determine the actual
retention time an oil soluble dye tracer
was applied in the field. The results
showed a peak concentration after 1-1/2
hours, and zero concentration of the
tracer after three hours. Therefore, the
3 hour actual divided by the 50 hour
calculated results in an actual 6%
hydraulic efficiency.
THE FOCUS SHIFTS TO WATER
QUALITY
With the advent of large scale
waterfloods in late 1940s, water quality
grew in importance. By 1960 water
quality was in the forefront of the minds
of all who dealt with water injectivity as
an enhanced oil recovery mechanism, or
simply for underground disposal... By
1970 the first Clean Water Act became
law in the USA.
Just as this Act
mandated cleaner water, it also became
a model for cleaner water for many
countries globally.
From the 1970s through the mid-1980s
a great deal of thought was given to
improving the inefficiencies of older
skim tanks and remnant oil from water

separation systems to improve water

quality.
Most were originally simply
empty tanks with no internals to aid in
distribution or separation, like the one in
Figure 4.

TYPICAL SKIM TANK

FIGURE 4

Field tracer surveys proved that shortcircuit flow paths do exist in nearly all
older skim tanks, regardless of design.
Retention times were documented at
less than 6% of the calculated
(expected) retention time, in test after
test. Many new concepts were tried in
an effort to improve water quality and
lower costs. Additionally, internal baffle
adaptations were also tried in attempts
to improve effluent water quality.
In this brief period the industrys
financial condition was unusually strong.
The value of crude was high enough
that every effort was made to capture
every last drop and send it to the
pipeline/refinery.
Therefore, more
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 16

Myths versus Realities

testing
of
higher
degrees
sophistication were funded.

of

New tank internals were developed,

installed, and performance tested.
Practical, proven test methods were
used. These include the application of
known quantities of fluorescene or
urinene dyes. But in general, the results
were quite dismal.
More and more test results proved that
most of these new designs still had low
hydraulic efficiency, and in many cases
even poorer separation. It seemed that
adding baffles simply caused more
mixing, and more re-entrainment rather
than achieving the expected increases in
separation efficiency.
IMPORTANT LESSONS LEARNED
As the boom years came to an end,
even more exotic designs were tried,
and reached as high as 21% hydraulic
efficiency. But, for the most part, they
proved to be too costly or too
unreliable. Furthermore, the poorest of
the new designs were proven to have
hydraulic efficiencies less than 0.1%.

fluid to a rate greater than the

separation velocity of the separable fluid
fraction. Obviously, when the horizontal
flow rate is greater than the vertical
separation flow rate, separation ceases
and mixing occurs.
The industry began to come to grips
with the fact that good flow
characteristics enhance separation as
much as increased separation time.
Another point learned was that

eliminating acceleration points, where

re-entrainment of separated fluids can
occur, was even more important than
vessel size. An example of this is a

vertical baffle or group of vertical baffles

with holes drilled in it/them. The holes
distribute the flow across the cross
section of the baffle and prevent short
circuiting by creating a usually small
pressure drop. However, the result of
this or any restriction is to create an
accelerated velocity. The fluid flow rate
must increase through each hole. As
was explained above, when the
horizontal flow rate due to this
acceleration is greater than the vertical
separation flow rate, separation ceases

However, in all of this, important new

lessons were learned. One of the more
important lessons was that increased

retention time alone does not

necessarily enhance the separation of
immiscible fluids. The reason for this is

that most of the various methods of

increasing retention time, such as
vertical baffles forcing a serpentine flow
pattern through a skim tank, increase
the horizontal flowing velocity of the

Eddy Currents in Orifice

Flow
FIGURE 5

2013 HTC, Inc.

This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 17

Myths versus Realities

and mixing occurs, as in Figure 5.

Furthermore, flow path studies have
proven over and over again that altering
flow direction can cause re-entrainment
of separable fluids. As the flow of a
fluid exits any office, the pressure drop
produces eddies which wrap around
back toward the orifice, increasing the
flow velocity even more, shearing and
mixing larger droplets into smaller
droplets and re-entraining dissimilar
fluids.
Obviously, these forces are
detrimental to separation.
HWSB SKIM TANK-GUNBARREL
DEVELOPED,
TESTED
AND
PATENTED
Then in the 1980s, the oil boom came to
an end. The price of oil fell back into
the teen, and finally into single digits
again. The strong financial position of
the industry vanished almost overnight.
Knowledge took a back seat to survival.
The lessons learned, and a large portion
of the knowledge gained, was lost as a
result of lay-offs and early retirement
programs.
One survivor of this inevitable downturn
was the desire by HTCs principal to
develop an efficient and simple gravity
flow oil-water clarifier that improves
performance to more reasonable levels.
Work continued through this period of
time, though on a scaled back level.
New designs slowly evolved, were tried,
and fine-tuned. Each gave way to the
next until the design known as the
HWSB was fully developed.

Finally, in 1991 HTCs principal was

granted Patent Number 5,073,266 on a
system that accomplishes the goal of
clarifying
produced
water
to
injection/disposal water quality! This
was followed by Patent Number
2,053,326 in 1996.
The new
technologies were trademarked and
labeled HWSB". Every HWSB is
engineered and designed for each
individual application under the watchful
eye of the named patent holder.

HWSB Skim Tank

Patent Drawing
FIGURE 6

The HWSB was a wide departure from

all previous conventional wisdom. Its
design is quite unusual. It has no
moving parts. It applies concepts not
found elsewhere in oilfield processing.
It looks outside our industry for
methods proven sound in other types of
fluid handling. It clarifies water. It
approaches the ideal condition of piston
displacement, resulting in very efficient,
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 18

Myths versus Realities

repeatable, and forgiving separation

efficiencies.
By altering the standard design slightly,
the HWSB becomes a crude oil
dehydrator (a modern gunbarrel/wash
tank) as well as an ultra-efficient water
and oil clarifier. It is the 21st centurys
ideal skim tank. And, where suspended
solids like iron sulfide or oil coated
formation fines, create interface layers
or settle to bottom further proven
additions to the basic design control
these conditions so they no longer affect
water or oil quality. When the standard
HWSB is modified with proven solids
and interface removal systems, cleaning
and normal maintenance can also be
prolonged indefinitely.

process in any way (no shut-downs,

nothing to de-pressure or drain!). This
high-tech improvement has been well
extremely received by all lease
operators, providing more operating
flexibility to put more oil in the sales
tank whenever possible, and cleaner
water in the water tanks!
THE HWSB DESIGN

The correct and proper design of each

HWSB process vessel is arduous and
quite time consuming. The inlet fluid
velocity MUST be slowed to less than
rise and fall rates of the contaminants in
the water phase. This must be done
carefully, and in stages, so the oil
separates upward to the top of the
vessel, and the solids separate
downward to bottom. Changing the
The operation of HWSB Skim Tanks
was further improved
direction of flow is
with the hydraulically
necessary in this
engineered water leg;
process,
so
another industry first.
avoiding shearing
HTC designed water
and
mixing
legs provide for low
velocities is critical.
pressure drop and
The annular spaces
large spillover weir
created by the
areas to maintain the
tank walls and
most uniform and
internal spreaders
stable
oil-water
are carefully sized
interface inside the
to accomplish this
HWSB,
where
up
to
the
stable
levels
are
maximum
flow
rating
of
each
critical
to
HWSB Skim Tank-Gunbarrel
performance. Then,
HWSB system.
FIGURE 7
HTC
added
an
externally adjustable
Over 90% of the
oil separation of larger oil droplets
water leg design that allows the
operator to manage the HWSB
occurs in the upper inlet area above the
interface levels without affecting the
inlet spreader.
The larger droplet
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

Myths versus Realities

normally account for a small fraction of

the total recoverable oil, except in
unusual cases when the inlet stream is
mostly oil. The remaining remnant oil
exists in such small droplets that they
remain suspended in the rapidly moving
water until its velocity is slowed. This
happens as the bulk water flow turns
downward and passes through in the
upper annulus area between the large
upper baffle and the shell of the tank.

August 1, 2013
Page 19

otherwise surely be re-entrained.

Each HWSB is designed for its specific
application. Each application specific
HWSB design manages all fluid flows
and velocities to maximize gas-liquids,
water-oil, and water-solids separation.
The
overall
hydraulic
efficiency
approaches 72%, a far cry from the 3%
efficiency of the typical gunbarrel.
Seventy-two percent is point at which
the maximum threshold of separation
breaks over to become mixing.
At
beyond 72% mixing energy takes over
and reverses the separation process
altogether.

As the water passes through the upper

annulus an eddy current is created
under the upper baffle by the pressure
drop through the annulus. This eddy
current pulls the downward flowing
THE NEWEST HWSB DESIGN
water under the upper spreader where
it redistributes across the entire cross
section and into a near-perfect plug flow
As the current oil boom propelled the
(piston displacement) flow
industry north into colder
pattern, with 100% of the
climate areas HTC designed,
water
distributed
into
developed, tested, proved, and
100% of the tank cross
patented a new Cold Weather
section.
The uniform
HWSB with the emphasis on
distribution
slows
the
freeze protection. This was
down velocity of the bulk
perceived to be a valuable
water enough so that
addition for the large Bakken
smaller oil droplets can
Field in Norht Dakota, and for
now rise at a rate greater
all of other very cold weather
than
the
downward
producing
horizons
where
velocity of the water and
wintertime operations demand
separate.
As these
a focus on freeze protection.
smaller oil droplets rise
they are collected under
HTCs innovative Cold Weather
HTCs Cold
the upper baffle and flow
HWSB
was
successfully
Weather HWBB
directly into the oil layer
tested in the winter of 2010Patent Drawing
through a dedicated pipe
2011.
It exceeded HTCs
Oil Dehydrator
conduit without having to
expectations,
and
the
Figure
8
flow through the more
expectations of the first end
turbulent inlet area again,
user. A patent application was
where
they
would
prepared and filed in April
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

nks,
oots

Separation In Oilfield Production Operations

Myths versus Realities

2012, and the Patent 8,496,740 was

granted to HTC on July 30, 2013.
21ST CENTURY SKIM TANKS

Three HWSB Skim Tanks

Figure 9
Today, over four hundred HWSB Skim
Tank (The 21st Century Gunbarrel)
systems are in service in the domestic
oilfields of the United States. Dozens
more are working to clarify oil and water
in the worldwide oil industry. Each and
every one of these was designed for its
specific application. And to date, they
have outperformed the expectations of
their owners.
ST

21

CENTURY GUNBARRELS

Not all applications focus on high water

cuts. Some applications still have a
need to dehydrate large quantities of oil
with smaller concentrations of water.
For these low water cut applications
HTC has developed a 21st century crude
oil dehydration gunbarrel. This unique
design focuses on maximizing crude oil
dehydration rather than water quality.
Oil with up to 30% water and emulsion
is the target process stream. Since

August 1, 2013
Page 20

many of these streams are larger in

volume, most HWBB Gunbarrels are
retrofits of existing larger API650 tanks,
or new applications in these same large
tanks. Smaller shop fabricated versions
are also common when produced oil
volumes are lower.
The HWBB uses a hydraulically
imbalances distribution system to meter
inlet fluid across the entire cross section
of
eh
Gunbarrel,
maximizing
dehydration
through
a
uniform
reduction in
upward
oil
velocity.
This allows
the heavier
water to flow
downward to
below the oil
layer where
it can be
collected and
withdrawn.
HTCs HWBB
Oil Dehydrator
Figure 10

Oil rises to
near the top
of
the
HWBB where it is collected throughout
360 of the tank, achieving uniform
collection of dehydrated crude oil to
maximize efficiency.
This High-Tech
design has proven itself to be more
efficient than conventional gunbarrels in
applications worldwide, ranging from
1500 BOPD to 250,000 BOPD.
HTC provides the entire engineering
design spectrum from entire facilities to
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.

Separation In Oilfield Production Operations

August 1, 2013
Page 21

Myths versus Realities

individual vessels the development of

AutoCAD PFD and P&ID drawings
through detailed civil and mechanical
engineering packages, all in AutoCAD
for oil and water separation/treating
systems.
ABOUT THE AUTHOR
Bill Ball has focused his
career on oilfield facilities
design for over fifty years.
During those five decades Bill
has received five patents and
has two more pending.
He has
accumulated a unique knowledge of
separation; a subject that he knows has
its roots in physics, hydraulics, and
chemistry. His hands-on experience has
pulled all of this together. Today, Bills
designs are used in hundreds of oilfield
facilities from SWD Plants in North
Dakota to waxy crude dehydration
facilities in India. Bill is founder and
president of High-Tech Consultants, Inc.
ABOUT HTC
HTC was incorporated in 1993 in Tulsa,
OK to provide various proprietary and
patented design technologies to the oil
and gas industry. These include the
patented warm and cold weather
HWSB Skim Tank (Gunbarrel), the
HWBB
Crude
Oil
Dehydrator,
proprietary desanding
flow-splitting
tanks and many others.
HTC is a
st
leading design firm for 21 century salt
water disposal (SWD) plants.
HTC is proud to be able to share this

paper with those interested in improving

their oilfield water-oil-gas separation
operations.
HTC is proudly responsible for every
individual HWSB design.
HTCs
attention to detail assures each and
every owner of an HTC HWSB, or any
of its other designs/technologies, of the
unprecedented success in each and
every application.

Three HWSB Skim Tanks

Flank Three HTC Polishing
Tanks in a SWD Plant
Figure 11

INNOVATION
HTC is innovative!
When a major
producer wanted to cut costs and help
build its own water clarification system
HTC designed this flotation cell. This
saved over $100,000, and months of
time!

Fiberglass Flotation Cell

Figure 12
2013 HTC, Inc.
This document was prepared by HTC
for its loyal clients and prospective clients worldwide.