2012 International Conference on Lightning Protection (ICLP), Vienna, Austria

Summary of 2011 Direct and Nearby Lightning Strikes to Launch Complex 39B, Kennedy Space

Center, Florida

C.T. Mata, A. G. Mata ESC

Kennedy Space Center, FL, USA Carlos.T.Mata@nasa.gov, Angel.G.Mata@nasa.gov

Abstract— A Lightning Protection System (LPS) was designed and built at Launch Complex 39B (LC39B), at the Kennedy Space Center (KSC), Florida in 2009. This LPS was instrumented with comprehensive meteorological and lightning data acquisition systems that were deployed from late 2010 until mid 2011. The first direct strikes to the LPS were recorded in March of 2011, when a limited number of sensors had been activated. The lightning instrumentation system detected a total of 70 nearby strokes and 19 direct strokes to the LPS, in 2011, 2 of the 19 direct strokes to the LPS had two simultaneous ground attachment points (in both instances one channel terminated on the LPS and the other on the nearby ground). In addition to the 70 nearby strokes, some more distant nearby strokes where captured on video records for which limited data were acquired. Instrumentation deployment chronological milestones, a summary of lightning strikes (direct and nearby), high speed video frames, downconductor currents, and dH/dt and dE/dt typical waveforms for direct and nearby strokes are presented in this paper.

Keywords-Lightning Protection System; LC39B; KSC, Data Acquisition System, Lightning Instrumentation System.

I. INTRODUCTION The installation of the meteorological and lightning

instrumentation systems at Launch Complex 39B (LC39B) started in February 2011 and March 2011 and were completed in September 2011 and May 2011, respectively. The triggering thresholds of the lightning instrumentation were adjusted based on previous test runs at the International Center for Lightning Research and Testing, (ICLRT), Camp Blanding, Florida, in which the same instrumentation was tested in 2009 and 2010. During the previous to last Space Shuttle mission, STS-134, the triggering thresholds of the system were reduced in an attempt to make the LC39B lightning instrumentation trigger from strikes at Launch Complex 39A (LC39A), where Endeavor was being prepared to launch. The thresholds were lowered from March 10, 2011, until shortly after the STS-134 launch on May 16, 2011. A few storms passed through the launch pad areas and many images of nearby strikes were captured with some limited number of usable waveforms. Note that the dynamic range of the measurements was not changed, so the majority of the waveforms for distant nearby strikes had very poor resolution. Shortly before the roll out of the last Space Shuttle

mission, Atlantis, STS-135, May 31 2011, a trigger signal from the LC39A lightning instrumentation system was connected as an input trigger to the LC39B lightning instrumentation system (June 21, 2011), in an attempt to increase the LC39B lightning instrumentation triggering capabilities and reduce the number of triggers from nearby events to both launch pads. This trigger signal from LC39A to LC39B was left in place even after Atlantis launched in July 8, 2011. The lightning instrumentation system at LC39B helped determine the location of LC39A nearby strikes during some storms prior to launch and even a strike within the LC39A perimeter the day before launch, preventing a potential launch scrub that would have delayed the mission and resulted in additional costs.

II. INSTRUMENTATION OVERVIEW AND SETUP Reference [1] and [2] describe the LC39B weather

instrumentation, which is composed of two data acquisition (DAQ) systems, Lightning and Meteorological, both running 24/7. The lightning protection system (LPS) consists of a catenary wire system (at about 181 meters above ground level) with nine downconductors connected to ground, supported by three insulators installed atop three towers in a triangular configuration. Field sensors’ signals are digitized and transmitted, via fiber optics, to centralized transient recorders located in the Pad Termination Connection Room (PTCR). All the LPS instrumentation, including the digitizers, transient recorders, and computers, are battery backed up.

A. Lightning Instrumentation This is an event driven, 100 megasamples per second

sampling system, which includes six synchronized high-speed video cameras installed atop each of the LPS towers (two per tower) and one remote high-speed video camera installed about five kilometers southwest of LC39B atop the Vehicle Assembly Building (VAB)). There are a total of nine current sensors installed at ground level on each of the nine downconductors; four dH/dt, 3-axis measurement stations; and five dE/dt stations composed of two antennas each. The electromagnetic sensors are all installed at ground level with elevations ranging from 5.9 to 7.4 meters height referred to the North American Vertical Datum of 1988 (NAVD88). All these signals are recorded using three synchronized transient

recorders (HBM GENESIS 16t). Fig. 1 shows an overview of the lightning instrumentation system installed at LC39B.

Figure 1. Panoramic view of LC39B, at the KSC, Florida, showing the lightning protection system and the architecture of the lightning

instrumentation system.

Window triggering and an OR gate are used to trigger the LC39B lightning instrumentation system from any of the thirty one ground level sensors (9 downconductor currents, 12 dH/dt, 10 dE/dt). Additionally a TTL signal from the LC39A lightning instrumentation system is also an input to the OR gate. After a qualified trigger is received, the signal of all the ground level sensors is recorded on a 30 millisecond time window with 50% pre-trigger sampling at 100 megasamples per second. The seven high speed video cameras record up to 455 frames with 312.5 microsecond exposure (up to 142.1 milliseconds of video at 3,200 frames per second) with a 1280x800 resolution, also with 50% pre-trigger. The VAB camera has been operational since spring of 2012.

B. Meteorological Instrumentation The meteorological instrumentation is a 24/7, 1 sample per

second, system, which provides air temperature, relative humidity, wind direction, and wind speed at four different altitudes (40.2, 78.3, 116.4 and 139.3 meters), on each of the four levels (A, B, C, and D) of the LC39B LPS towers. In addition, there are two rain gauge stations (at ground level) that measure rain precipitation and accumulation, adding for a total of 26 meteorological sensors. Fig. 2 shows an overview of the meteorological instrumentation system architecture installed at LC39B.

The meteorological data is displayed on the KSC internal television system (OTV), showing the one-minute running average of all the measurements in addition to the maximum wind gust and wind direction over the last minute. Additionally, the meteorological data will be transmitted (fall of 2012) to the 45th Weather Squadron (WS), Cape Canaveral Air Force Station (CCAFS), Florida, to support weather related launch decisions.

 

Figure 2. Panoramic view of the LC39B, at the KSC, Florida, showing the architecture of the LC39B meteorological instrumentation depicting a typical

lightning protection tower and the sensors’ locations.

III. CHRONOLOGICAL MILESTONES The weather instrumentation (lightning and meteorological)

of LC39B has been incrementally increasing after the LC39B LPS was built. Table I shows the milestones of the deployment of these DAQ systems.

TABLE I. LIST OF CHRONOLOGICAL MILESTONES OF THE LC39B WEATHER INSTRUMENTATION SYSTEM FROM ITS CONCEPTUAL DESIGN.

Action Time Conceptual Design 2007 Construction 2008-2009 Instrumentation Enclosure fabrication, testing and installation 2010

DAQ installed January 2011 All (4) Tower 1 meteorological stations active February 2011 All (9) downconductors instrumentation, towers 2 & 3 meteorological stations (8) and 2 (out of 6) high speed video cameras active

March 2011

First and last nearby lightning strike recorded (2011) Mar. 30th & Oct. 10th

First and last direct lightning strike recorded (2011) Mar. 31th & Aug. 14th

4 (out of 6) high speed camera and all (12) magnetic field measurements active April 2011

All (6) high speed video cameras and all (10) electric field measurementsa active May 2011

Meteorological data shown on OTV, additional trigger signal from LC39A and additional (temporary) high speed video camera in Launch Control Center (LCC)

June 2011

1 (out of 2) rain stations active August 2011 All (2) rain stations active September 2011 VAB camera installation March 2012 Meteorological data sent to 45th WS (expected) October 2012

a. From May until July there were some electric field measurements troubleshooting, so, limited dE/dt measurements are available for this period.

IV. ACQUIRED LIGHTNING DATA The lightning instrumentation system was active from mid-

March of 2011 and it triggered on 14 different days acquiring a total of 48 lightning flashes with 89 return strokes. These flashes are organized between direct and nearby. The latter are also grouped between close nearby (within LC39B perimeter) and nearby, referring to all the strikes with attachment points outside the LC39B perimeter. Fig. 3 shows colored tear drops, yellow (for negative) and blue (for positive) to indicate the locations of all the direct and close nearby strikes acquired by the LC39B LPS during 2011.

Figure 3. Satellite view of the LC39B indicating the location of the three lightning towers, 1-3. Yellow lines show the projection to ground of the catenary wire system and downconductors. The red circle delimits a 1

kilometer square area centered in the pad. Colored tear drops, yellow for negative and blue for positive, indicate the location of only the direct and

close nearby strikes to the LC39B, additionally the strike location for the two return strokes with multiple simultaneous attachment points is shown outside

of the LC39B perimeter.

A. Direct: The LPS has been struck directly by 8 negative lightning flashes with a total of 19 return strokes: • 5 flashes attached to the towers.

• 3 flashes attached to the catenary wire system or downconductors.

• 3 single-stroke flashes.

• 5 multiple-stroke flashes (minimum of 2 return strokes and maximum of 8 return strokes).

• Only one multiple-stroke (3) flash had all subsequent return strokes terminating at the same location as the first return stroke.

• 2 return strokes had two simultaneous attachment points, direct (tower) and nearby outside the pad perimeter.

B. Close nearby (within the LC39B perimeter fence): There were 3 nearby lightning flashes, with a total of 6 return strokes, striking within the LC39B perimeter: • 1 four-stroke flash striking the perimeter fence (all

three subsequent strokes following the path of the first stroke).

• 2 single-stroke flashes (one negative and one positive) terminating on the ground.

C. Nearby: The lightning instrumentation system captured 40 nearby (outside the LC39B perimeter fence) lightning flashes with a total of 64 return strokes: • 2 nearby subsequent strokes, whose first strokes

attached to the LPS.

• 26 presumed single-stroke flashes (due to the limited range of the LC9B lightning instrumentation system it is not possible to label these as single-stroke flashes).

• 14 multiple-stroke flashes (minimum of 2 return strokes and maximum of 6 return strokes and). Note that due to the limited range of the LC39B lightning instrumentation system it is likely some of the subsequent strokes were not captured.

V. STATISTICS From late March until December of 2011, the LC39B was

struck by downward lightning strikes as observed in high speed cameras videos. A total of 11 flashes terminated within the LC39B perimeter fence (out of 48 flashes detected by the LC39B lightning instrumentation system, or about 23%) which has an area of approximately 1 km2, see Fig. 3. About 55% of these flashes (6 of 11) had multiple return strokes and 33% of the multiple-stroke flashes (2 of 6) had all the return strokes (4 and 3) terminating at the same location.

One of these flashes was a positive (single-stroke) flash representing 9% (1 of 11) of all the flashes detected by the LC39B lightning instrumentation system within an area of 1 km2. Consequently, the estimated ground flash density for LC39B during 2011 was 11 flashes/km2/year with 9% of positive flashes. Furthermore, the 2011 estimated ground stroke density for LC39B was 25 strokes/km2/year with 4% positive strokes. Within this area, about 73% of the flashes (8 of 11) and 76% of the strokes (19 of 25) attached directly to the LPS and 18% (2 of 11) of these flashes had multiple simultaneous attachment points, with one branch attaching directly to one of the LPS towers and a second branch attaching to ground (outside of the LC39B perimeter).

These results seem to contradict the predictions of the LC39B LPS Monte Carlo simulations previously run [1] and [3], where over a period of 10,000 years, using a ground flash density of 17 flashes/km2/year (with 5% of positive flashes), 26% of the flashes struck the LPS and 74% struck ground in the vicinity of the LC39B within an area of approximately 1 km2.

About 63% of the direct flashes were multiple-stroke flashes (5 of 8) and only 20% of them (1 of 5) had all of its return strokes (3) terminating at the same location.

Each of the three towers of the LC39B LPS was struck directly by lightning at least once, with tower 1 being impacted by 2 flashes (10 return strokes), tower 2 by 4 flashes (4 return strokes) and tower 3 by the first return stroke (with the largest measured peak current of 2011) of a three-stroke flash with the 2nd return stroke striking tower 2 and the ground (simultaneously) and the 3rd return stroke striking ground at the same location as the 2nd return stroke. Tower 2 was the only tower struck twice with direct strikes and simultaneous nearby strikes outside the LC39B perimeter fence (about 560 and 615 meters east and south-east of tower 2).

From the downconductors measured currents to ground, considering only direct strikes to the LPS:

1. The median peak current, calculated as the algebraic sum of downconductor currents to ground, was 29.1 kA for 16 return strokes (excluding a direct stroke which saturated two downconductor measurements and strokes with simultaneous ground attachment points).

2. Rise-times, 10% to 90%, were between 1 and 6 µs, with an arithmetic mean of 2.9 µs.

3. The minimum, arithmetic mean and maximum inter stroke time intervals from:

• 11 subsequent return strokes (direct strikes to the LPS) were 23, 84 and 180 ms.

• 14 subsequent return strokes (direct and nearby strikes within the LC39B perimeter, about 1 km2 area) were 17, 71 and 180 ms.

• 38 subsequent return strokes (direct and nearby strikes detected by LC39B LPS) were 1.5, 116 and 389 ms.

VI. SELECTED WAVEFORMS A representative set of waveforms comprised of the LC39B

LPS 9 downconductor currents, 12 dH/dt, 10 dE/dt are included for two selected cases, direct and nearby strikes to the LC39B LPS. Additionally, salient data from the U.S. National Lightning Detection Network (NLDN) and the Cloud-to-Ground Surveillance System (CGLSS) is presented for these events [4] and [5].

A. Direct Lightning Strikes Recorded downconductors’ currents show a consistent

incident current division mainly into two portions defined by 1) the shortest electrical distance between the strike location to ground and 2) the rest of the LPS. Such current division is depicted in Fig. 4 where a direct strike to tower 1 (see Fig. 5) is illustrated. The measured downconductors’ currents are shown in Fig. 6 and in this case, the shortest electrical distance between the lightning strike and ground is through downconductors 1 and 2, which are connected to the air strike

terminal of tower 1 and provide a short electrical path to ground. The other portion of the incident current flows through the rest of the LPS to ground, also with a dependency on the shortest electrical distance of each of the remaining downconductors. The largest measured currents to ground, as expected, flow through the downconductors closer to the strike location.

Figure 4. LC39B LPS current distribution scheme due to a direct strike (to tower 1). Colored arrows (red and blue), indicate the division of the incident

current into two main groups.

Figure 5. First return stroke of a negative lightning three-stroke flash striking tower 1 of the LC39B LPS on August 14 of 2012 at 21:10:15.7870656 (UTC).

Rate of change of the magnetic and electric fields measured at LC39B are shown in Fig. 7 and Fig. 8, respectively. Larger rate of change of fields are also observed on the closer stations to the strike location.

A direct comparison between the LC39B LPS calculated current to ground and strike location with the information provided by CGLSS and NLDN, for a direct strike to Tower 1 on 8/14/2011 at 21:10:15.787065 UTC, reveals the following:

1. The LC39B lightning instrumentation measured a peak current to ground of -54.4 kA, 10%-90% rise-time of 6.04 µs, transferring 7.75 Coulombs to ground in 1 ms.

2. CGLSS reported a peak current of -34.1 kA (37% peak current underestimation) with a location error of 68 meters southwest of tower 1.

3. NLDN reported a peak current of -39.8 kA (27% peak current underestimation) with a location error of 700 meters southeast of tower 1.

Figure 6. LC39B LPS downconductors measured currents to ground for the direct strike to tower 1 on August 14 of 2011 at 21:10:15.7870656 (UTC).

Figure 7. LC39B measured dH/dt for the direct lightning strike to tower 1 on August 14 of 2011 at 21:10:15.7870656 (UTC).

Figure 8. LC39B measured dE/dt for the direct lightning strike to tower 1 on August 14 of 2011 at 21:10:15.7870656 (UTC).

During 2011, it was found that both systems, CGLSS and NLDN, underestimate the peak current by 10-40% compared to the LC39B LPS measured currents to ground. Additionally the minimum, mean and maximum absolute location errors calculated for all the (19) direct strokes to the LC39B LPS were 51, 190 and 467 meters for CGLSS (12 samples) and 133, 572 and 1,714 meters for NLDN (14 samples) [4] and “unpublished” [5].

It is worth noting the availability of meteorological data, not included in this manuscript, for all the recorded lightning events of the LC39B LPS during 2011.

B. Nearby Lightning Strikes For nearby strikes, the current distribution among the LPS

downconductors is conceptually different than for direct strikes and it can be seen as currents flowing from the strike location into the LPS through the closest downconductors and then leaving the LPS through the downconductors farther away from the strike location. The current distribution can be visualized by defining three groups of downconductors identified by an imaginary line that goes perpendicular to the azimuth of the strike location, see Fig. 9.

The first group is made of all the downconductors between the strike and the imaginary line, which allows current to flow into the LPS. The second group is made of all the downconductors past the imaginary line, which allows the current to flow out of the LPS. The third, and less obvious

group, is made of the downconductor(s) within close proximity to the imaginary line, in which no significant current flow is observed. A high speed video camera frame, corresponding to the return stroke of nearby strike presented in this section is shown in Fig. 10.

Figure 9. LC39B LPS current flow scheme due to a nearby strike. Colored arrows (red, blue and green) indicate the three different current flow groups

and brown dashed lines illustrate the imaginary lines indicating the azituth of the lightning strike location and its perpendicular line referenced to the center

of the path.

Even though a thorough analysis of nearby strikes has not been completed at the present time, corresponding downconductors’ current data show a tendency to balance the flow of current for both, previously referred, first and second groups. The arithmetic sum of all currents to ground result in

the displacement current through the LPS. As expected, closer nearby strikes seem to result on currents to ground with higher frequency components (see Fig. 11) than farther away nearby strikes.

Figure 10. Nearby single-stroke negative flash striking withing the LC39B perimeter fence on August 14 of 2012 at 21:05:44.0800717 (UTC).

Similar to the direct strike, the larger rate of change of the magnetic and electric fields are observed closer to the strike location, see Fig. 12 and Fig 13.

A direct comparison between the LC39B lightning instrumentation system detected strike location and the information provided by CGLSS and NLDN reveal that for the nearby negative single-stroke flash striking 473 meters North-Northwest of the LC39B center (within the LC39B perimeter fence) on 8/14/2011 at 21:05:44.080012 UTC, CGLSS reported strike location error was 122 meters NLDN reported strike location error was 364 meters. The peak current reported by both, CGLSS and NLDN for this event were -15.8 kA and -18.4 kA, respectively. Note that the LC39B lightning instrumentation automated capability to estimate the nearby stroke peak currents is still under development and it is expected to be completed by 2013.

Figure 11. LC39B LPS downconductors measured currents to ground for the nearby lightning strike within the pad perimeter on August 14 of 2011 at 21:05:44.0800717 (UTC).

Figure 12. LC39B LPS measured dH/dt for the nearby lightning strike within the pad perimeter on August 14 of 2011 at 21:05:44.0800717 (UTC).

Figure 13. LC39B LPS measured dE/dt for the nearby lightning strike within the pad perimeter on August 14 of 2011 at 21:05:44.0800717 (UTC).

VII. CONCLUSIONS After less than one year of deployment, the new lightning

instrumentation system at LC39B has shown significant improvements over the previous lightning instrumentation system used at KSC. The instrumentation system has proven to be extremely reliable and accurate, capturing a series of astonishing images and set of waveforms for typical and unusual lightning strikes. There is an increased interest to potentially deploy similar systems throughout KSC and CCAFS, in which case, in the future, there may be a comprehensive an integrated lightning observation system with synchronized lightning instrumentation over an area of several square kilometers where ground-truth data will be acquired.

REFERENCES

[1] C.T. Mata and V.A. Rakov, “Evaluation of lightning incidence to elelments of a complex structure: a Monte Carlo approach,” International Conference on Grounding and Earthning & 3rd International Conference on Lightning Physics and Effects (Ground 2008; 3rd LPS), Florianopolis, Brazil, November 2008.

[2] C.T. Mata, V.A. Rakov, T. Bonilla, A.G. Mata, E. Navedo and G.P. Snyder, “A new comprehensive lightning instrumentation system for PAD 39B at the Kennedy Space Center, Florida” International Conference on Lightning Protection 2010, Cagliari, Italy, September 2010.

[3] C.T. Mata and V.A. Rakov, “Monte Carlo modeling of lightning incidence to structures” International Workshop on Electromagnetic Radiation from Lightning to Tall Structures, Montreal, Canada, July 2009.

[4] C.T. Mata, A.G. Mata, V.A. Rakov, A. Nag and J. Saul, “Evaluation of the performance characteristics of CGLSS II and U.S. NLDN using ground-truth data from Launch Complex 39B, Kennedy Space Center, Florida” International Lightning Detection Conference (ILDC), Broomfield, USA, April 2012.

[5] C.T. Mata, A.G. Mata, V.A. Rakov, A. Nag and J. Saul, “Evaluation of the performance characteristics of CGLSS II and U.S. NLDN using ground-truth data from Launch Complex 39B, Kennedy Space Center, Florida” International Conference on Lightning Protection (ICLP), Vienna, Austria, September 2012, in press.