Tuesday, April 4, 2017

We will be covering the topic of lightning protection this week.  We'll spend today discussing the protection of structures.  Then we'll look at protecting electronics systems from transients created by lightning.  We'll also look briefly at personal lightning protection. 

Most of this lecture and the next are adapted from material in Uman's The Art and Science of Lightning Protection (Cambridge Univ. Press, 2008; paperback version 2010).  An interesting survey of the costs of various types of lightning damage can be found on the first 3 or 4 pages in Chapter 2.  Here are are few examples of the interesting facts and statistics found there:

The 1977 lightning-caused power blackout in New York City cost about $350 million.  30% of all power failures are due to lightning with an annual cost close to $1 billion.

 

Estimates of the costs of lightning damage to insurance companies range from 1/3 to $1 billion.  In total about 5% of insurance claims involve lightning (50% of the claims in Florida in the summer involve lightning)

 

30,000 house fires are caused every year by lightning.  About half of the 20,000 wildfires every year are caused by lightning.  The cost of fighting wildfires was about $1.5 billion in 2006 (a record year).

 

Lightning damage to commercial airlines costs ~$2 billion per year.  Damage to military aircraft is comparable.

 

Lightning causes damage to perhaps 100,000 computers/year.




You will not learn enough details and specific information to be able to go into business installing lightning protection on homes and commercial buildings in Tucson.  To find that kind of information you need to consult documents such as the
NFPA 780: Standard for the Installation of Lightning Protection Systems issued by the National Fire Protection Association (it is available for purchase ($61.50) but can be read for free online).  In Europe the International Electrotechnical Commission is the relevant agency I believe.  They have issued IEC 62305 Protection against Lightning also available for purchase (1046 Swiss Francs.  There is roughly a 1:1 correspondence between Swiss Francs and US Dollars so that is many times more expensive).

I do hope that you will look for and better understand and appreciate some of the lightning protection that has been installed on buildings in Tucson and on the UA campus (mainly on the newer buildings).


Chances of being struck
We'll use my rental house as an example (the landlord's house is located next door and was struck by lightning a few years ago).  Here is a quick calculation of the chances that the house might be struck.  We'll multiply the effective area of the house by the lightning flash density for Tucson.

The dimension of the house are roughly: 20 m long (l) by 12 meters wide (w).  The top of the pitched roof (h) is about 5 m high.  To determine the effective area we must extend the actual horizontal dimensions of the house by 2 times the height (2 x h) on all four sides of the house.


Then we multiply this area by the strike density from the last lecture.



You would need to question whether it is worth the expense of installing lightning protection (especially since I rent and don't actually own the house).  Though you don't really need a direct strike to cause damage especially to sensitive electronics inside the house.

You can do the same kind of calculation for a normal power distribution line.  We'll assume the arms on the power pole that hold the power lines are 5 m wide and are about 10 meters above the ground.  The effective area of 100 km of line then would be



 

In this case some lightning protection is probably advisable.  Often a ground wire will be run above the current carrying wires.  Lightning then will hopefully strike the ground wire (and be safely carried to ground) instead of the current carrying wires.




Protection system components
Lightning protection for a house or building usually consists of 3 components:
an air terminal (lightning rod or Franklin rod), a down conductor, and grounding.


 


Dimension A in both figures is 20 or 25 feet and is the maximum spacing recommended by the NFPA between lightning rods.  Dimension B in the lower figure is 20 feet and is the maximum spacing between rods located along the edge of the roof.  Two down conductors at opposite corners of the roof are shown in the top figure.  The bottom figure has a down conductor at each of the buildings corners.


Solid, low resistance electrical connections are needed in lightning protection systems.  The web site for Thompson Lightning Protection Inc. gives you a pretty good idea of the various components that might be used.

Zone of protection

For structures less than 50 feet tall

Aa lightning rod provides a 45o or 60o cone of protection.  The 45o cone from a single lightning rod is shown at top in the figure below.



A larger structure can be protected with multiple rods with overlapping zones of protection.


A different approach is used for larger structures.  We need to recall that the striking distance, d,  is the distance between the leader tip and the ground at the moment upper connecting discharges are initiated.




Anything that intrudes into the sphere of radius d is at risk of being struck.

      

In the rolling sphere method, you imagine rolling a sphere of radius d (d is the striking distance) over the structure being evaluated for lightning vulnerability.  Parts of the building touched by the sphere are at risk; those parts of the building could be struck.  Portions that aren't touched are safe.



The height of the building at left is less than or equal to d.  The sides of the building aren't touched by the sphere, they've been shaded blue and are safe.  The sphere does touch the top of the building, it is vulnerable.  Lightning protection would need to be installed on the roof.  The building at right is quite a bit higher than d.  Much of the sides as well as the roof can be struck by lightning.  Protection would need to be installed on the top sides of the building and the roof.  Fragments of building materials knocked off the sides of buildings by lightning are apparently a serious hazard to people on the street below.


Air terminals could be mounted on the top of a tall buildings.  Additionally, the building designers could make metal window shades and window frames part of the lightning protection system by connecting them to the down conductor.


Note the spacing of the air terminals must be such that the rolling sphere doesn't touch the roof in between adjacent lightning rods.


What value of d should be used?  Researchers have tried to relate the striking distance to the amplitude of the peak return stroke that follows.  Generally an expression of the form

d = A (Ipeak)b


is used.  An example curve is shown below




Recommended values for d for different levels of protection are given in the table below (based on Table 4.1 in Uman's "The Art and Science of Lightning Protection").

protection level
d
minimum peak I
% rs>Imin
I (highest level)
20 m
3 kA
99%
II
30
5
97
III
45
10
91
IV (lowest level)
60
16
84
 
Level I is the highest degree of protection.  With a striking distance of 20 m, all return strokes with a peak current greater than 3 kA should striking the lightning protection system.  This level would provide protection against 99% of return strokes.  The NFPA recommends a striking distance of 30 m or 46 m. 

Interestingly it is the small amplitude return stroke that are likely to pass in between air terminals and strike a protected building.  We don't really know how common these small return strokes are because they are below the trigger threshold of most detection systems.
   


Protection of a roof

A roof can be protected by multiple lightning rods.  An old question is whether pointed rods or blunt rods work best.  Some recent research (Moore et al. (2000a and 2000b))  suggest that the blunt rods might be more effective.  The NFPA allows either type to be used.  Rods should be at least 10 inches tall.


So called "nonconventional" arrays of points that try to prevent formation of an upward connecting discharge (dissipation arrays) or rods that use radioactivity or high voltage to attempt to initiate an "earlier" upward connecting discharge (thus offering a larger zone of protection) are generally not thought to be any more effective than conventional "Franklin" rods (see Uman and Rakov (2002) for an clear and interesting discussion of this topic).


A wire mesh covering the roof of a building or vulnerable installation on the ground can be used instead of lightning rods.  A wire mesh covered the administration trailer at the rocket triggered lightning site in Florida for example.


The following table gives recommendations for the spacing of the wires in a wire mesh (based on Table 4.2 in Uman's "The Art and Science of Lightning Protection")


protection level
d
mesh spacing
I
20 m
5 m x 5 m
II
30
10 x 10
III
45
15 x 15
IV
60
20 x 20


A metal roof also provides good protection from lightning (connected to down conductors and grounding).  The NFPA recommends that the roof be at least 3/16 inch thick.






Overhead wires are used to protect launch installations at the Kennedy Space Center Launch of an Atlas V rocket at Launch Complex 41 Cape Canaveral Air Force Station (source of this image)

The insulated support in the left lower figure is on the structure and not directly overhead the launch vehicle.

Down conductors 

Air terminals should be connected to as many down conductors as possible.  Lightning rods on a house should be connected to at least two down conductors (at opposite corners of the house for example).

 
There are several reasons for this:

 

(i)  The down conductor will have some resistance and impedance.  Connecting inductors or resistors in parallel reduces the combined impedance.



(ii) T
he B fields from two down conductors positioned on opposite sides of a building will point in opposite directions in the interior of the building.  This will reduce the strength of the B fields inside the structure and reduce the intensity of voltages and currents induced in electronic circuits inside the building.

 

The recommended minimum crossectional area for down conductors is about 50 mm2 for copper  (that correcponds to a radius of 4 mm, a diameter of 8 mm or roughly 1/3 inch)

 

Almost closed loops on the down conductor should be avoided.  This adds impedance and sparks are likely to jump across adjacent points on the down conductor.


Rather than going around a protrusion on a building wall, it would be better to go through the protrusion and keep the down conductor as straight as possible.

If a lightning protection system is struck it can easily rise to a potential of a few megavolts (see the discussion at the beginning of Lecture 16).  The down conductor should be "bonded" to (electrically connected to) metallic objects within about 5 m to prevent sparking from the down conductor to nearby metallic objects.  In the figure below (my rental house again) the down conductor has been connected to a water faucet at (1), to a vent pipe at (2), and to an evaporative cooler at (3).  A second down conductor extends from the roof to ground at Point 4.



Here are typical fields needed for spark initiation and propagation:

3 x 106 V/m is need to breakdown air at sea level
500 kV/m is the average field needed for propagation of negative polarity leaders
300 kV/m is the average field needed for positive leader propagation.


Grounding
Grounding rods are generally copper clad steel and should be at least 8 feet long.  The bottom end should be at least 10 feet deep.


Grounding can be improved by connecting grounding rods to a counterpoise, a loop of grounding wire that encircles the structure.  This reduces potential differences inside the loop.


The grounding can be also be connected to a buried mesh under the structure or to interconnected rebar in the concrete base of the structure.  This is well illustrated in an article  "Lightning Protection of Structures and Personal Safety" by V.A. Rakov (2000).


Here is a table of resistivities of various materials (Table 5.1 in Uman's "The Art and Science of Lightning Protection").

Material
Resistivity in ohm-meters
ocean water
0.1 - 0.5
ground, well & spring water
10 - 150
lake & river water
100 - 400
rain water
800 - 1300
commercial distilled water
1000 - 4000
chemically clean water
250000
clay
25 - 70
sandy clay
40 - 300
peat, marsh & cultivated soil
50 - 250
sand
1000 -3000
moraine
1000 - 10000
ose (calcereous remains)
3000 - 30000
 

Resistivity is not the same thing as grounding resistance.  We'll do a relatively straightforward calculation of grounding resistance.  The rather unusual geometry is shown below and will make the calculation more manageable.  A current I is injected into a hemispherical piece of metal that is buried in soil. 



We'll assume that the current spreads out evenly in the soil.  We divide I by the area of half a sphere to determine the current density J.


Now that we have an expression for E, we can determine the potential.



We'll determine the potential difference (voltage) between a point deep in the soil (r = b) and a point close to the grounding electrode (r = a).



Here's the remainder of the calculation.  A grounding resistance of 40 ohms would be pretty good for Tucson.


Computing the grounding resistance of grounding rods or buried cables would be more difficult.  So we'll just cite two additional tables (Tables 5.2 and 5.3) from Uman's book.  These give ground resistances of a 1/2 diameter vertical grounding rod and a 1/2 inch diameter buried horizontal wire as a function of conductor length and soil resistivity (100 ohm-meter is relatively a relatively soil conductivity, 1000 ohm meter is relatively poorly conducting soil).

rod length
resistivity
100 ohm-meter
resistivity
1000 ohm-meter
3 m
35 ohms
350 ohms
6
22
220
9
15
150


wire length
resistivity
100 ohm-meter
resistivity
1000 ohm-meter
50 m
4.0 ohms
40 ohms
100
2.6
26
200
1.4
14



References: 

C.B. Moore, G.D. Aulich, and W. Rison, "Measurements of Lightning Rod Responses to Nearby Strikes," Geophys. Res. Lett., 27, 1487-1490, 2000a


C.B. Moore, W. Rison, J. Mathis, G. Aulich, "Lightning Rod Improvement Studies," , J. Appl. Meteorol., 39, 593-609, 2000b

V.A. Rakov, "Lightning Protection of Structures and Personal Safety," Intl. Lightning Detection Conf., Tucson, 2000.

M.A. Uman and V.A. Rakov, "A Critical Review of Nonconventional Approaches to Lightning Protection," Bull. Am. Meteoro. Soc., Dec. 2002