Experimental Arrangements To Study Lightning Attachment Characteristics in Northern Australia

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

EXPERIMENTAL ARRANGEMENTS TO STUDY LIGHTNING

ATTACHMENT CHARACTERISTICS IN NORTHERN AUSTRALIA

J.R. Gumley, F. D’Alessandro, M.A. Austin

ERICO Lightning Technologies, AUSTRALIA.

Abstract: This paper describes the new experimental in 1997 to allow resources to be utilised at the new
arrangements established in 1996 and 1997 for the sites.
1997-8 storm season in Northern Australia and some
of the preliminary data that have been recorded.
Research has been directed toward comparison of air
terminal performance under natural conditions
through measurement of emission currents in the
milliseconds before a return stroke. One recent case is
shown in detail where the ground stroke termination
point is estimated to have been less than 500 m from
the test terminals. The recorded data show initial
electric field pauses and steps in the approaching
leader of approximately 50 µs, with a reduction to less
than 10 µs in the 30 µs preceding the return stroke.
Comparison of these new results with previous
measurements suggests that the field intensification
produced by the present air terminal arrangement may
be too high for optimum leader development, and that
some of the test terminals should be reconfigured.

1. Introduction

Darwin is located on Latitude 14o South in an area


where tropical monsoons form from November
through to March each year. Storms traditionally form (a)
South East of Darwin in the Humpty Doo area and
progress toward the North West. In the past, a site at
Darwin airport was used on a passive basis with six
air terminals on 9 m masts. Emission currents were
digitised at 2 MHz for a period of 3 ms before and 1
ms after the return stroke. The pulse and continuing
escalation currents associated with non-intercepting
upward leaders from this site has been previously
reported by Gumley [1].

Since that time, two new sites have been established.


One site was established in 1996 at Kowandi, some 30
km south of Darwin and close to Humpty Doo (see
Figure 1). The other site was established in 1997 and
is located near the coast on the Casuarina campus of
the Northern Territory University. The air terminal (b)
arrangement at this site is a replica of the Kowandi
site but using the digitiser equipment previously Figure 1: Kowandi site. (a) Air terminal arrangement
located at the airport. The airport site was dismantled (identical at Casuarina). (b) Air conditioned container
housing the instrumentation.
The general level of activity during the wet season in season we will be removing the resistive shunts. The
Darwin is 80-100 thunderstorm days per year. advantage of this change is that the number of
Digitiser triggering from activity near the Darwin channels recorded will be halved, giving us the option
airport site during 1992-1997 shows that there are of doubling our sampling rate.
approximately 5, 8 and 8 thunderstorm days in
December, January and February respectively. The
first season at Kowandi (1996-7) produced triggering
days of 6, 6 and 3 respectively. During the 1997-8
season, the corresponding figures for Kowandi were 8,
6 and 11 and for Casuarina they were 7, 4 and 9. The
radius over which triggering is achieved appears to be
in the order of 5 km. Trigger counts on any active day TO PC BASED
DIGITISER SYSTEM

can vary from 2 up to 100 events.

Apart from the (unmanned) measurement of air


terminal emission currents, the field research program
for 1997-8 included the pre-empting of discharges by Figure 2: Simplified schematic of the differential
rocket triggering, with wire spooling commencing measurement system using current transformers.
150m above the test site, and the use of an Automatic
Lightning Discharge Progressing Feature Observation
System (ALPS) brought to Australia by Gifu Four air terminals were used in the field tests, namely
University researchers. The latter was successfully a spherical air terminal (diameter 350 mm) which
trialed at Kowandi in 1996 with a late afternoon, operates on a capacitive coupling principle (AT1), a
daylight, recording of light emissions associated with prototype ellipsoidal air terminal (500 mm diameter),
a progressing downleader. Analysis of the ALPS data also based on capacitive coupling (AT2), a grounded
is only just beginning, but the results will be available sphere the same size as AT1 (AT3) and a Franklin
for publication in the next ICLP. rod (AT4) with a conical tip culminating in a
hemispherical cap approximately 3 mm in diameter.
The long term objectives of these continuing efforts in The specific details of the physical arrangement and
the field testing of air terminals are: (i) to assess their the electrical (analogue and digital) components of the
performance over a range of blunt, sharp and active field experiment for both sites are summarised in
configurations; (ii) assess the role of wind in the Table 1.
changing of Franklin rod performance by corona
(space charge) displacement; (iii) observe the electric At Kowandi, the triggering signal of the ALPS and
field necessary to initiate and maintain upward leader the 20-channel digitiser were interlinked to allow time
propagation under natural conditions; and (iv) to comparison of emission currents with the upward
record the propagation velocities of up leaders, a leader progress velocity being signified by light
parameter critical to the modelling of lightning emissions.
protection systems.

3. Preliminary results
2. Technical details
A total of more than 500 trigger events were recorded
Apart from physical air terminal layout, a number of from both sites during the 1997-8 storm season in
other changes have been implemented at the two new Darwin. The largest event was recorded at the
sites. The main changes have occurred in the Casuarina site on March 1, 1998. The electric field at
analogue measurement system. The air terminal ground level at the time of the return stroke was
emission current measurements are now made using a greater than 100 kV/m. Anecdotal evidence as well as
differential system employing (isolated) current simple downleader modelling indicates that the strike
transformers (CT). The basic system is illustrated in point may have been within a radius of 500 m of the
Figure 2. This sensing system was combined with the air terminals. The magnitude of the electric field
existing resistive shunt system so that comparisons resulted in significant current pulsing in all the air
could be made in the event that erroneous results were terminals in the array. Hence, this event shall be used
obtained because of the ground loop effects inherent to examine the response of the air terminals.
in the ground referenced shunt sensing. The CT
system was so successful that for the coming lightning
Table 1: Summary of technical details relevant to both sites.

Location Kowandi & Casuarina, Darwin


Number and type of air terminal Four - short-tip Dynasphere, 500 mm ellipsoid , grounded sphere
and Franklin rod respectively
Mast height and separation 15 m and 20 m respectively
Mast configuration/orientation Linear; SW-NE at both Kowandi and Casuarina
Prevailing storms From the SE
Instrumentation:
• Sensing system Resistive shunts (43 mΩ) and high frequency pulse current
transformers
• Recording hardware Coaxial cables ~ 200 m long carrying sensed voltages back to
GageScope ADC cards installed in a PC running off-the-shelf
software
• Channels recorded Kowandi - 12 out of 20 available channels (1 x shunt, 1 x CT signal
for each air terminal, 2:1 (46 kV/m FSD) & 10:1(230 kV/m FSD) e-
field, e-field slope, ALPS trigger);
Casuarina - 7 of 8 channels (e-field as above plus shunt & CT for
AT 1 & AT2) on PC1 (ex-Darwin Airport) and similar for PC2 (ex-
Florida) but for AT3 & AT4
• Sampling rate/data window 2 MHz with a 4 ms window divided 3:1 around the return stroke
• FSD emission current Shunts - 50 A; CT - 0.1 V/A, i.e., FSD = 10 A
Electric field sensor:
• Type Induction probe
• Triggering principle dE/dt, hence triggering the hardware on the collapse of the e-field at
the return stroke
• Sensitivity ~23 kV/m per volt of output
Other equipment Video camera(s) at both sites
Kowandi - ALPS and full set of high powered rockets (single use and
RMS) and associated equipment for artificially triggering lightning

The true magnitude of the field at ground level may 30 µs, there is a scattering and irregularity in the
have been higher because the location of the sensor (a electric field which may be associated with the final
fast field antenna) was such that it was not completely jump. The rapid increase in stepping frequency
outside the “electrostatic shadow” of the air terminal suggests an increase in leader velocity upon its
masts. Hence, we can assume the value of 100 kV/m approach to the strike point. This observation tends to
is a lower limit. Computer modelling of the electric confirm the work of Yokoyama et al [2] who obtained
fields around the test arrangement has shown that the a visual (ALPS) recordings of the velocity in the final
gross field (ignoring local microgeometrical features) stages of leader approach.
existing at the top of the terminals at the time of the
return stroke was 50-100 times the field at ground Figure 3(a) shows the initial, significant current
level. If local features are taken into account, e.g., the pulsing behaviour of the Franklin rod. This activity
sharp tip on the Franklin rod, the not-so-sharp tip on commenced almost 300 µs before the return stroke at
the spherical air terminal etc, then the local electric a time when the electric field strength on the ground
fields are at least one order of magnitude higher. This was ~ 25 kV/m. The pulse magnitude is ~ 5 A and the
indicates that, under static thunderstorm conditions of pulsation period is ~ 20 µs. Figure 3(b) shows the
5-10 kV/m, the sharpest terminal(s) would have been current pulsing behaviour of the Franklin rod just
emitting significant corona. before the return stroke. Here, the electric field on the
ground is 60-100 kV/m, the pulse magnitude is
Examination of the initial recording of electric field greater than 10 A and the pulsation period is ~ 4 µs.
shows the expected pausing of the downleader for
periods of approximately 50 µs. However, in the final During the 300 µs prior to the return stroke, it is
stages, this pausing and the interpause interval estimated that the Franklin rod pumped more than
decreases to approximately 10 µs. During the final 500 µC of charge into the air above its tip. A
simplified calculation using Poisson’s equation shows
that this amount of charge can significantly reduce the
electric field at the tip of the rod. This reduction can
result in unfavourable conditions for leader inception.
The observed > 10 A pulses indicate that streamers
had thermalised but could not progress due to low
field strength ahead and so were collapsing, making
matters worse for the next streamer because of the
space charge left behind.

Interesting deductions can be made when the present


data are compared with some obtained from a
similarly strong event at the Darwin airport site in Figure 4: Emission current measurement for large
1994 (some results were presented in [1]). The sphere, made on January 26, 1994, at the Darwin
emission current trace shown in Figure 4 was airport site. The vertical line denoted with a “T”
obtained from a test terminal comprising a floating marks the point in time of the return stroke. FSD
sphere, of diameter 750 mm, surrounding a central current is 20 A.
grounded rod. The configuration was such that the
spherical surface was grounded under high static
fields but isolated during the approach of a The difference is striking. In this case, there is very
downleader. The overall height was 9 metres. little pulsing until the last ~ 50 µs, at which time
Computer modelling shows that the field there is also a significant and continuous interpulse
enhancement factor of this arrangement is emission current that eventually increases to ~ 20 A,
approximately 13, i.e., up to two orders of magnitude while pulse current magnitudes exceed 10 A. The
lower than the terminals in the present test site. magnitude and nature of these emissions suggest that
an upward leader was both created and progressing.

4. Conclusions

Comparison of the results in Figures 3 & 4 suggests


that the field enhancement factors (Ki) of the
terminals in the present arrangement (Kowandi and
Casuarina sites) are too large, resulting in a space
charge which causes the near field to be reduced to a
point below that required to sustain an upward leader.
(a) This result tends to support the theoretical studies
presented elsewhere in these conference proceedings
(see papers by Gumley & Berger, D’Alessandro &
Gumley). Continuing work in this area will see a re-
definition of the terminal parameters. In particular, it
appears that the Ki value can be optimised to
maximise the probability of upward leader initiation.

References

[1] Gumley, J.R.: “Lightning interception and the


(b)
upleader”, 22nd ICLP, Budapest, Hungary, paper
R 2-11, 1994.
Figure 3: Current pulsing observed in the Franklin
rod during the March 1, 1998 event recorded at the [2] Yokoyama, S., Miyake, K., Suzuki, T. & Kanao,
Casuarina site. The FSD current is 10 A and FSD S.: “Winter lightning on Japan sea coast -
electric field (lower trace) is 46 kV/m. (a) Initial Development of measuring system on progressing
current pulsing (to ≈ −300 µs, relative to the return feature of lightning discharge”, IEEE Trans.
stroke). (b) Current pulsing in the 30 µs immediately Pow. Del., Vol. 5, No. 3, pp. 1418-1425, 1990.
prior to the return stroke.

You might also like