As explained in the chapter "Ultrasonics" elsewhere in
this description, it is the traveltime of an emitted sonar pulse from
the emitter point to the target on the wall and back that is measured
in sonar surveys. If the distance of the path traveled is to be
determined then it is necessary to know precisely what the velocity is
in the medium the pulse has traveled through.
The majority
of sonar systems in use worldwide contain no means for determining the
acoustic velocity. Instead the velocity is estimated from tables and
from experience. For instance in brine caverns, in which the acoustic
velocity lies between 1740 m/s and 1870 m/s depending on the density,
the velocity is assumed to be 1800 m/s. The maximum distance error in
the radius is proportional to the inaccuracy of the acoustic velocity
(-60 m/s to +70 m/s), which corresponds to an error of -3.33% to
+3.88%. As this error is squared when considering the area and volume,
the error range increases to -11.09% to +15.05%. |
| Sonar tools of the type EDO WESTERN are equipped with an
integrated acoustic velocity module which works on the "swing-around
principle". In the swing-around principle an acoustic emitter transmits
acoustic pulses - in this case with a frequency of 1 MHz - in a slow
and continuous sequence. These acoustic pulses are diverted when they
strike two surfaces at 45° and subsequently arrive at a separate
receiver transducer. The electronics connected to this receiver
transducer amplify the acoustic pulse and transmit it again from the
emitter transducer. A continuous oscillation results, the duration of
which corresponds to the distance covered. System traveltimes in the
electronics and any phase errors that arise cannot be taken into
account. The system operates generally with a resolution of 5 m/s. Due
to the encapsulated design of the tool there is only a mediocre
exchange of medium so it is not possible to adequately prevent the
build up of unwanted deposits, which can considerably affect the length
of the distance. The reference distance, a critical factor for the
accuracy, is only about 20 cm. Doppler effects resulting from the
tool's own movement during a log run would be compensated by the upward
and downward signal path. However, this system is not intended for use
during log runs; its purpose is to determine acoustic velocity values
for individual survey depths. |
Acoustic velocity module based on the swing-around-principle |
The SOCON BSE acoustic velocity module is made up of a
compensated, multi-stage system with difference compensation. It is a
modular part of the BSE tools. Adapted acoustic pulses are emitted into
the medium in programmable transmission and reception phases from a
sonic emitter and receiver transducer. The acoustic pulses strike two
spatially offset reflectors, which create echo responses.
In order to ensure a good exchange of medium in the vicinity of the
measuring length, the system has an open design, which allows the
medium to freely flow around it. A drawback of such systems is that
deposits invariably form on the reflectors, which would falsify the
readings. To overcome this drawback only the lower sides of these
reflectors are used. i.e. the acoustic pulses are transmitted from the
bottom to the top. Every response includes the outward and return path
and so excludes doppler effects caused during survey runs.
Not only the two traveltimes to the reflectors (tshort and tlong) are
of use, but also the difference of the two traveltimes (tdiff) can be
used as the traveltime between the reflectors. |
Microprocessors installed in the tools as well as software
routines calculate the acoustic velocity for each of the three
reference lengths and subsequently perform an error analysis. The error
components caused by the electronics, such as system delays, affect
only the "long" and "short" paths and not the difference. All that
remains is a system-independent error that can contain phase errors
owing to the interpretation method; these are compensated by
microprocessors and software. Differences or deviations between the
tools are taken into consideration by the software. The software does
this by requesting the serial number and characteristics of each sensor
and then applies an appropriate calibration table. This technique is
used for all sensors fitted in SOCON tools.
To achieve the
greatest possible accuracy, the measuring length is 50 cm and operates
with a time resolution of 10 ns. For an acoustic velocity of 1800 m/s,
for instance, and a traveltime of 277 ms for tlong, the traveltimes are
sufficiently long to obtain exact and reproducible data. The system is
characterized by a resolution of < 0.1 m/s and a reproducible
accuracy of < 0.4 m/s. Continuous survey runs throughout the entire
cavern and in all media - regardless of whether gas, oil or brine - are
not only a part, but a requirement of surveying. |
SOCON BSE acoustic velocity module with difference compensation |
Despite adhering to the highest quality standards, some
influences cannot be properly detected and taken into account (a chain
is only as strong as its weakest link). So, assuming we want to keep
the effort and expense at a reasonable level, we have to accept that
the acoustic velocity can be determined only in the survey axis. This
means that the measuring accuracy is affected in media with an
inhomogeneous content (e.g. brine concentration) between the survey
axis and the cavern wall.
In brine caverns the
concentration increases towards the cavern wall. This fact causes the
propagation velocity of the acoustic pulse to increase as it approaches
the cavern wall, which results in a shorter time being measured, and
consequently the distance appears too short.
This effect can be corrected by running in the cavern a PROFIL LOG with
the transducer oriented horizontally and subsequently a PROFIL LOG with
the transducer tilted at 45°. In this way different distances are
obtained for one point on the cavern wall. From this data, compensation
values can be derived and a distance-dependent correction performed. |
Ray diffraction and apparent shortening of the traveltime caused by an increase in acoustic velocity towards the cavern wall |
| Experience over several decades has shown that generally a
substantial increase in concentration takes place only in the last few
meters next to the cavern wall, which means the error remains
relatively small even without compensation. Nevertheless, it must be
borne in mind that especially in brine-filled caverns for which no
compensation has been carried out, the values measured are too small,
in general by approximately 4 millimeters per meter. |
| The increase in concentration also affects
measured values in another respect, namely through the deviation of a
beam emitted from a tilted tool head. The tilted beam is diffracted
from the assumed straight beam path by an amount depending on the
increase in concentration. This results in survey values in the roof
area being recorded from a steeper beam. The roof therefore appears to
hang down. |
Deviation of rays caused by changes in the brine concentration indicate, if not corrected, a "hanging roof" |
On the other hand in the floor area the beam is diffracted
to become flatter, which results in an apparent raising of the floor.
For this reason it should be attempted to optimize the tilt depths and
the associated tilt angles.
The investigations of a customer in a
cavern with a flat roof (blanket was air) showed some interesting
results. Subsequent to sonar surveying the brine was emptied and the
roof resurveyed with mining surveying techniques. The results indicated
that the roof "hanged down" by 0.3 meters at a distance of 60 meters.
To increase the accuracy in such cases it is necessary to apply a
correction value.
The above mentioned problems do not occur in gas and oil caverns. In
oil caverns, however, attention must be paid to layering of different
oil types, whereas in gas caverns the density and consequently the
pressure and the acoustic velocity increase linearly with increasing
depth. |
By recording logs, such as acoustic
velocity, pressure, CCL and temperature, in combination with a sonar
survey, it is possible to recognize and appropriately deal with problem
zones.
Information gained from the log recordings is used
to optimize the survey procedure and adapt it specifically to the job
to be carried out. |
|
