Climatological, Meteorological and CFD Analysis

CIGRE 2012 21, rue d’Artois, F-75008 PARIS http : //www.cigre.org B2_207

Effects from downbursts on overhead lines; Meteorological thunderstorm study – climatological, meteorological and CFD analysis

C. KARNER S. TSCHANNETT A. BECK H. LUGSCHITZ M. RATHEISER H. KAUFMANN 1) Austrian Power Grid APG W. GEPP Zentralanstalt für Meteorologie Austria H. KAUFMANN und Geodynamik ZAMG Weatherpark GmbH Austria Austria 1) current affiliation

SUMMARY

In February and March 2008 cyclone “Emma” crossed some countries of central Europe and caused severe damages. In very local areas thunderstorms with downbursts were embedded in “Emma” and led to collapses of overhead lines.

Downbursts are downward winds with high velocity within a thunderstorm concerning a relative small area. They occur seldom but can have severe effects and are responsible for various local damages on buildings, forests and infrastructure as e.g. overhead lines. In many cases in history, such damages were not brought into connection with downbursts. The Cigre reports 350 and 410 deal with localized high intensity winds. They state that “…more research is needed to study the interaction of localized winds with supports...” The question arose if the terrain triggered wind speed-up effects. Another question was which wind forces and directions occurred during these events. To investigate the influences of high wind speeds on damaged overhead lines, a pilot study was carried out. Three dimensional wind fields where computed and analyzed regarding to speed-up effects. This paper presents the investigations, climatological aspects as well as CFD case studies (Computational Fluid Dynamics) and shows the amplification factors for the wind speed.

The method which was developed in the project can be used for any other region to simulate tendencies of wind speed-up, regardless if they come from downbursts or not. It helps to a better understanding of the reason for wind induced failures.

KEYWORDS Downburst, downdraft, tower failures, cyclone, high intensity winds

[email protected]

Damages on Overhead Lines by cyclones Emma and Vivian

From February 29th to March 2nd in 2008 the extratropical cyclone Emma crossed North and Eastern Europe and caused serious damage especially in Germany, Switzerland, Austria, Czech Republic, Hungary and Slovakia. In very local areas thunderstorms with downbursts were embedded in Emma and led to severe damages on houses, forests, all kind of facilities and to collapses of overhead lines. In APG´s grid in Austria towers of a 110kV and a 220kV line failed in a very small area of app 1 km in diameter, whilst another 110kV line 600m apart remained more or less unaffected. In February 1990 the cyclone Vivian caused similar but less damages in the neighbouring area. This was not brought into connection with a downburst at that time.

Figure 1: area with damages from cyclone Emma. The two blue lines indicate the area where the overhead line towers collapsed. [1]

In the Cigre publication 350 a downburst is defined as “…a strong convective down-draft inducing an outward flow of damaging winds when reaching the ground. The downdraft makes contact with the ground and then spreads outwards, causing severe winds at low altitudes. These events are often associated with thunderstorms“. It is there also mentioned: “Downdrafts can sometimes be larger than tornadoes in extent, i.e. more than one span can be affected by an event.” [2]

During Emma at the 220kV line seven suspension towers and at the 110kV line five suspension towers collapsed. No foundations failed. There is no indication for poor material quality, neither for the steel angles nor for the bolts. No indication for brittle fracture was found. Conductors and towers were not covered by ice. The conductors did not brake due to the high content of steel (single conductor aluminium/steel 340/ 110 mm2), damages of the conductors in the clamps were found. The lines were built 1958 and 1979. Both lines were designed for loads exceeding the minimum values given in the relevant Austrian standards. The design ice loads were of 60–70 N per meter conductor. The material quality of all components was in order and was not the reason for the collapses.

Eyewitnesses of Emma reported heavy rain and hail during the disastrous wind, which lasted only a few minutes. Nobody was hurt, no property was damaged from broken towers. In Hungary and in the Czech Republic Emma created similar damages on lines from 110kV to 380kV. These faults occurred at several locations.

Figure 3: no brittle fractures of tower steel at Figure 2: typical situation of a damaged tower. No failures of foundations. the tower´s angles (built 1958 resp. 1979.)

Damages on buildings, forests and other facilities were enormous. The situations were reported as natural catastrophes. In APG´s grid with 12.000 towers the collapses caused by Vivian and Emma were the only ones from effects of heavy wind so far.

Climatologic situation

Twice within 19 years the region of St. Peter am Hart in Austria was hit by windstorms which caused severe damage. The region is mainly characterized by two large flow pat- terns, cyclones associated with westerly winds and weather fronts and anticyclones in association with easterly winds and more stable weather conditions. According to the Köppen-Geiger Classification the climate is “warm temperate” and “fully humid” with “warm summers” [3]. Most precipitation is detected during the summer. Usually higher mean wind speeds occur in winter and spring. Cyclone Emma passed Central Europe with two severe gale events, the first stronger one passed on March 1st (figure 4) and the second on March 2nd. The highest measured gusts in Austrian flat terrain occurred on March 1st and reached about 140 km/h in the city of Salz- burg [4]. At meteorological sites in mountainous areas occurred 140 -165 km/h.

Figure 4 : Reanalysis of sea level pressure from March 1st 2008 showing extratropical cyclone Emma (labelled with T) over Sweden and the associated cold front over South and Eastern Europe (solid blue line) and the so-called convergence line associated with thunderstorms (dashed blue line). [5]

Figure 5: Radar echoes of precipitation associated with cyclone Emma, corresponding to the convergence line shown in figure 4 approa- ching the region of St. Peter in 2008, red lines 380kV OHL, green lines 220kV OHL (radar picture: Austro Control)

The meteorological analysis takes into account observational informa- tion from all meteorological sites in Austria as well as sophisticated in- terpolation algorithms. Nevertheless the area of the affected overhead lines does not agree with the areas of the highest analysed gust speeds. This fact suggests the hypothesis that very local effects, such as downbursts, are responsible for the collapse of the towers. Indeed, thunderstorms with precipitation and hailstorm were embedded in Emma (see figure 5) and in very local areas also downbursts occurred [1]. At Ranshofen (about 5 km distance to the area of collapsed towers), the highest measured gusts lay at about 103 km/h which normally occurs at this measuring site every 7 to 10 years. By the time of this highest wind gusts the temperature dropped about 8°C within 10 minutes (see figure 6). According to radar-echoes and the character of damages on lines and other damages in the adjacent area it was found out, that the wind speed must have exceeded the measured ones at Ranshofen [1]. It was concluded that a downburst occurred with most likely wind speeds of 210-220 km/h at the area of the collapsed lines, in the level of standard measuring height 10 m above ground level.

Figure 6: Time series for measured data at meteorological site Ranshofen on March 1st

2008 during Emma (time in UTC) based on 10 minute measures. Direction mean wind dd

[1/10°], direction wind gusts ddx [1/10°],mean wind speed ff [km/h], speed wind gusts

ffx [km/h], rain sum rr [1/10 mm], Temperature tl [°C], pressure p [hPa]. At the time of

the highest wind gust (09:40 UTC) the temperature dropped abruptly, pressure increased and in association with the thunderstorm heavy rain occurred (within one hour more than 7 mm were detected).

At the end of February 1990 extratropical cyclone Vivian affected great parts of Europe with high winds and resulting damage to buildings and forests [6]. Just a few days later secondary extratropical cyclone Wiebke developed and passed Central Europe. In Austria highest wind speeds were measured on February 26th (Vivian) and March 1st (Wiebke). In the concerned region - at Linz-Hörsching (298 m above MSL) - about 155 km/h were measured on February 26th (which here occurs averagely about one time every 100 years). Associated with the second windstorm Wiebke wind speeds of 162 km/h were measured at Linz-Hörsching which is the highest value ever measured there. The front passage of Vivian was associated with a strong drop of surface pressure and temperature and accom- panied by strong precipitation amounts [7]. At neighbouring sites in Germany maximum wind speeds on February 26th reached 133 km/h in Passau and 115 km/h in Mühldorf. At the mountain site Feuerkogel (1618 m above MSL) in Upper Austria, gusts of about 169 km/h were detected, and on March 1st 180 km/h.

Research program and methodology

In view of the observed wind storms a model-based climatological investigation has been carried out to assess the spatial pattern and temporal distribution of strong wind events in Austria. A comprehensive model-based climatology has been put together. This synthetic climatology covers the time period 1974 to 2008 and is based on atmospheric reanalysis- data downscaled to 1x1 km using a chain of numerical weather prediction models. Observational data is taken into account implicitly on the synoptic scale (i.e. more than 100km) originating from the reanalysis-data and on a regional scale (typically 20x20km) to ensure that the simulated wind speeds are in agreement with the observed regional climatology. The synthetic climatology reflects the overall distribution of high wind speeds in the area. Major storm events are captured and the spatial pattern of individual events is simulated correctly. Despite the high-resolution, numerical weather prediction models are usually unable to simulate correctly the temporal and spatial location of individual thunderstorms and, in particular, the possibility of downbursts associated with severe thunderstorms. Therefore further investigations had to be carried out, to answer the following hypotheses: Either the local terrain is able to trigger the wind speed-up as e.g. explained in Cigre 410 [8] or downbursts associated with severe thunderstorms related to the convergence line (shown in Figure 4) are responsible for the collapse of the towers. To that aim, meteorological measurements and simulations with a high-resolution Computational Fluid Dynamics (CFD) model have been carried out.

Local Meteorological Measurements

The local wind conditions served as a basis for the wind flow studies. In addition to the existing climatological stations of ZAMG three wind measurement stations were operated by Weatherpark in order to get a higher resolved station net.

Figure 7 shows the distribution of wind direction at the meteorological measurement station Biburg which is the nearest one to the collapsed towers. It was operated for one year. The distribution shown is a combination from a 13 year wind climatology from the station Ranshofen and the 1 year wind data time series from the station Biburg (2009-2010) in order to get a wind climatology as input values for the computer simulations. The maximums at Westsouthwest and Eastnortheast are due to the large scale flow regime

(domination of westerly winds in the middle latitudes). The third maximum at Southeast is due to a local thermal wind system. The prevailing wind direction for strong winds which are mostly associated with frontal passages – is Westsouthwest.

Figure 7: Distribution of wind directions at the meteorological measurement station (Biburg) nearest to the collapsed towers.

Modelling the Wind Field by CFD Simulation

To investigate the influences of terrain and/or downbursts on the overhead lines, a high resolution Computational Fluid Dynamics (CFD) model was adapted to this task. Steady state simulations were carried out for four different scenarios: Thunderstorms within Emma and Vivian were simulated in a 2.4 x 2.4 x 0.7 km domain with and without downburst. There were two air flow inlets in the computational domain: one horizontal inlet in the “ceiling” of the domain which served as the inlet for the air from the downburst. The other, vertical inlet, was situated upstream of the transmission towers and let the ambient wind flow into the computational domain. Magnitude and direction of the ambient winds were taken from the climatology explained above. The downbursts in the model did not result from model physics, but were provided as boundary conditions at the horizontal inlet. What was calculated by the model were the modifications of the provided flow by the terrain (dark gray in figure 8) and other features like buildings and forests (light gray and green in figure 8, respectively).

Results Figures 8 and 9 show the results of one scenario (thunderstorm within cyclone Emma with downburst) from two viewpoints. The backward trajectories illustrate the path of the air that reaches the power lines (red circles). The legend in the figures shows amplification factors with respect to the downburst wind speed: factors greater than 1,0 mean amplification of wind speed, factors less than 1,0 mean deceleration. There is a vertical N-S cut coloured by amplification factors in figure 8 in addition to the backward trajectories. The location of the downburst is assumed to be west of the power lines (right hand side in figure 9, see arrow annotated with “D”) and has a wind speed at the inlet of 104 km/h. The mean flow preceding the downburst is about 30 km/h from Westnorthwest (see arrow annotated with “M” in figure 9).

The maximum amplification of the downburst air can be seen in the vertical cut in figure 8 (orange colours). The transmission lines lead directly into that maximum which is caused when the air hits the ground and is deflected upward again. The other viewpoint in figure 9 reveals how the accelerated air reaches the towers at a top factor of 2,0 (corresponds to a top speed of 208 km/h). Another nice feature is the horizontal mean flow which is whirled aloft by the downburst and deflected downwards (left hand side in figure 9). The vertical cut in figure 8 shows the amplification of the wind speed and an upward deflection of the air after it has reached the

6 ground. The CFD simulations of a thunderstorm in cyclone Vivian show similar results: a downburst wind speed of 119 km/h is amplified by a factor of 2,3 which results in a calculated top speed of 274 km/h at the towers (not shown in figures).

Figure 8 (above): Simulation results for scenario “Emma with down-bursts”. Backward trajectories and vertical N - S cut coloured by amplification factors. Terrain is dark gray, villages light gray and forests are green. Red circles show transmission towers, „M“ denotes the horizontal mean flow, „D“ the downburst. View direction: NW. Factors in the scale: amplification factors with respect to the downburst speed.

Figure 9: Same as figure 8 but with view direction SW. The amplification of the downburst air by factors of up to 2,0 results in top wind speeds of 208 km/h at the transmission towers. The highly turbulent and complex flow field suggests that the direction of the extreme strong winds must have varied in high frequency during the downburst event.

Conclusions

The results of the study led to the conclusion that the local topography contributes very little to an enhanced probability of the collapse of overhead lines near St. Peter. The damages to the lines from Emma and Vivian result from storms in combination with severe con- vection. The winds near the towers were both strong exceeding 200 km/h and very turbulent. The results of the CFD simulations show that within seconds upward and downward components from vortices induced near the ground changed on small areas and literally shook the towers until they collapsed. The exact local positions of the collapsed tower are arbitrary at the scale in focus. On a larger geographic scale the probability for such events is related to tracks of severe thunderstorms.

BIBLIOGRAPHY

[1] „Wissenschaftliche Analyse der Gewitterfallböen im Bereich von St. Peter am Hart am 1. März 2008“, Holzer A. and G. Pistotnik, ESSL 2008, Vienna [2] “ How overhead lines respond to localized high intensity winds”, Cigre Publication 350, 2008, Paris [3] “World map of Köppen-Geiger Climate Classification”, Kottek M., J.Grieser, C.Beck, B.Rudolf, and F. Rubel, updated Meteorologische Zeitschrift, Vol. 15, No. 3, 2006, Gebrüder Bornträger, Berlin, Stuttgart online available at http://koeppen-geiger.vu- wien.ac.at/pics/kottek_et_al_2006.gif [4] “Severe Weather Report” from Central Institute for Meteorology and Geodynamics (ZAMG http://zamg.ac.at/klima/klima_monat/unwetterbericht/?jahr=2008&monat=03 2008, Vienna [5] Reanalysis archive of Central Institute for Meteorology and Geodynamics (ZAMG), 2008, Vienna [6] „Stürme in Österreich im Zeitraum 1990 bis 2005“, Svabik O., internal severe weather archive of Central Institute for Meteorology and Geodynamics (ZAMG), 2007, Vienna [7] “Assessment of the Wind Gust Estimate Method in mesoscale modeling of storm events over West Germany”, Pinto J. G., C. G. Neuhaus, A. Krüger and M. Kerschgens, Meteorologische Zeitschrift, Vol. 18, No. 5, 2009, Gebrüder Bornträger, Berlin, Stuttgart [8] “Local wind speed-up on overhead lines for specific terrain features Working Group”, Cigre Publication 410, 2010, Paris