Gareth Evans STV. Peter Tabolt STV. Boston, MA. Philadelphia, PA. Simulation and Mitigation of Magnetic Field Created By Traction Power Systems

Transcription

1 Simulation and Mitigation of Magnetic Field Created By Traction Power Systems Peter Tabolt STV Vehicle Specialist Boston, MA Gareth Evans STV Senior Vehicle Specialist Philadelphia, PA 1

2 Presentation Purpose Explain effects of distortion Explain sources of distortion Explain prediction of distortion Explain methods of mitigation 2

3 Effects of Distortion Why does it matter? Many instruments are affected by very small disturbances in the ambient magnetic field e.g. MRI TEM SEM Instruments calibrated to ignore steady state ambient geomagnetic field approx 500mG Distortion in ambient field as small as 0.1 mg can cause noticeable distortion in the images obtained 3

4 Effects of Distortion Result is distortion in measured image 4

5 Sources of Distortion Geomagnetic Distortion Placing a ferrous mass in a magnetic field locally distorts the Earth s ambient magnetic field. 5

6 Sources of Distortion Light Rail Electrical System Feed current to LRV generates magnetic field Vertical Transverse Longitudinal 6

7 Sources of Distortion Vertical Currents (risers & pantograph) create a magnetic field in the longitudinal (B L ) and transverse (B T ) directions Horizontal Currents (contact wire, rails & feeder wires) create a magnetic field in the vertical (B V ) and transverse (B T ) directions Vertical Current B T B L Horizontal Current B V B T 7

8 Sources of Distortion Currents flow in a loop. A larger loop area will give rise to a larger magnetic field at a given distance from the loop. A larger current will give rise to a larger magnetic field. High Current e.g. from a high output voltage substation Low Current e.g. from a low output voltage substation Magnetic field R combine d3d 8

9 Mitigation of Distortion at Source Methods of mitigation Reduction of generated distortion Shielding at instrument 9

10 Reduction of Generated Distortion Reduce Line Current Below certain speeds, no effect on performance Appropriate for pedestrian areas 50% reduction in distortion. 10

11 Reduction of Generated Distortion Battery or Super Cap Power or Inductive or Switched Power Transfer to eliminate Traction Power Currents and Contact Wire Inductive Power Transfer System 11

12 Reduction of Generated Distortion Cutting Edge Technology: Mostly in Testing Phase Limits on Battery or Super Cap Distance especially with long grades Switched Contact Systems have problems with Snow and Ice Inductive Systems may have problems with track debris and ice Virtually Eliminates Traction Power Distortion Geomagnetic Remains 12

13 Reduction of Generated Distortion The more widely used Hi-Lo Power Distribution Most of the feed current flows in the feeders close to the return current except in the vicinity of the train. This cancels nearly all of the magnetic field away from the train and reduces it near the train by a factor of 8 to

14 Hi Lo Feeders Most widely used source mitigation method in US : Single Hi Lo Feeder geometry St Louis Metrolink Washington University since 2006 Double Hi Low feeder geometry Minneapolis for the Central Corridor Project Proposed for Maryland MTA Purple Line on University of Maryland Campus Seattle Sound Transit Background Papers Holmstrom, Reitter and Hamilton of LTK on the Seattle and Minneapolis Projects Zaffanella of Enertech on Saint Louis MTA Purple Line Paper developed by STV 14

15 Single vs. Double Hi Lo Feeders The single hi-lo feeder system runs Single feeder below the centerline of each track The double hi-lo feeder system 2 feeders, one below and outside each rail. Cancellation of the vertical fields is considerably better which is important for locations near the track. Which system to deploy depends on the equipment being protected and construction constraints. 15

16 Z Axis field milligauss Double Hi Lo vs. Single Hi Lo for Vertical Field Cancellation Vert Field One 5 car train passing observer followed by two in opposite directions Longitudinal Dis tance of N ear Track Train from Observer in Feet 100 feet from near track double hi lo 100 feet from track single hi lo 16

17 Y Component of B Field (mg) For transverse field, negligible difference One 5-car train passing followed by twotrains in opposite directions single hi lo double hi lo Lead Car Position on Track Relative to Observer in feet 17

18 Design Issues for Hi Lo feeders Cross bonding between tracks for traction return should be done only at the same point as the power feeders are linked between tracks (usually at the substation) Cross bonding tracks alone breaks up the current symmetry and greatly reduce attenuation. Design for midpoint of contact wire wear Monitoring and maintenance is required to maintain the design level of attenuation Feeder Connections Leakage of return current from the tracks to ground. 18

19 Prediction of Distortion 19 The University of Maryland required the MTA to provide a 3-D model showing how the main magnetic distortion was limited to a bubble around the train. The two of us together with Jalal Gohari of Parsons Brinkerhoff developed the following approach: Obtain typical LRV geomagnetic data from the Minneapolis Central Corridor Project Develop software to model the traction power magnetic distortion under various mitigation methods Add the two together with worst case addition point for point.

20 Prediction Part 1 Geomagnetic Distortion Geomagnetic Distortion Prediction based on extrapolation of measured data from the passage of a 3 car LRV train 20

21 RMS Disturbance (mg) Prediction of Distortion- Geomagnetic The magnitude of the resultant field and field distribution may be generated from the data 40 Geomagnetic Field Disturbance Generated by the passage of a 3 Car Train 35 25' from alignment ' 50' 75' 84' 100' ' from alignment 75' from alignment Time (s) 21

22 Field Distortion (mg) Value (mg) Prediction of Distortion- Geomagnetic The peak disturbance was plotted against distance from the train to make an estimate of an equation for the disturbance decay rate ' 35.09mG 50' 6.22mG y = 89407x Peak Geomagnetic Field Disturbance 75' 2.39mG 84' 1.728mG 100' 1.217mG Distance from car Peak Disturbance (mg) Disturbance Trend Line (mg) Predicted Geomagnetic Distortion The field decay may then be extrapolated and actually provides conservative estimate ' 10.34mG 55.00' 5.05mG ' 0.51mG ' 1.02mG 700.0' ' 0.01mG 0.10mG Distance From Alignment (Feet) Field Distortion (mg) 22

23 Prediction of Distortion - Traction Current Field By far, the largest part of the distortion is from the Traction Power Currents Unmitigated traction field R combine d3d 23

24 Prediction of Distortion - Traction Current Field It is easy to find software that will simulate current distribution of moving trains or to predict fields if the current is known. Combining is more difficult. So-- we wrote our own Mathcad Program. 24

25 Prediction of Distortion - Traction Current Field One vehicle moving between two substations 25 Used circuit equations to calculate the currents flowing in feeders, catenary, risers, and within the vehicle. Recalculated every unit distance as the vehicle moves down the track

26 Prediction of Distortion - Traction Current Field Biot Savart Law for a current segment Using the Biot Savart Law we calculated the magnetic field from each current segment. Using the superposition principles from circuit and field theory we were able to extrapolate to multiple cars and trains 26

27 Prediction of Distortion- Combined The geomagnetic field distortion is added to the field generated by the traction current This may be displayed as a 3d graph of the overall effect of the passage of the train. Level of Distortion 2d cartographic plane 27

28 Prediction of Distortion- Combined The footprint can be used to show the distortion as a train moves along the right of way Location of train Contour line for 0.1mG distortion 28

29 Prediction of Distortion - Traction Current Field The Mathcad program is parameterized to allow study of different variables such as: Number of cars or trains Direction of travel of each train Traction power feed geometry Height and longitudinal placement of measurement device Contact Wire Wear Mitigation employed Certain cases of return current leakage to ground 29

30 Prediction of Distortion- Combined A feature was also developed to find the worst case magnetic fields at each distance as trains moved past a point. The calculation result for each train position and distance from the alignment is stored in a matrix which was evaluated in a peak hold fashion. 30

31 Protection at the Equipment For very sensitive equipment, mitigation at the source may not be sufficient. Need Protection in the lab 31 (1) Courtesy TMC website for MagnetX Active Field Cancellation System (2) New Six-Layer Magnetically-Shielded Room for MEG D. Cohen1,2, U. Schläpfer3, S. Ahlfors1, M. Hämäläinen1,4, and E. Halgren1

32 Shielding Shielding is a common means of protecting against electromagnetic fields Difficult to do with quasi static fields produced by transit lines and equipment. Must divert the field rather than reflect or absorb. With triple layer Mu-metal rooms huge field reductions can be achieved but very expensive and hard to maintain. E.g. Martinos Meg Lab shielded room > $2 million. NMR machines can be purchased with extra Mu-Metal Shielding to reduce sensitivity and reduce the interference they create themselves. 32

33 Active Cancellation Active cancellation systems drive one or two Helmholtz coils on each axis to cancel magnetic field disturbances 33

34 Active Cancellation effectiveness 20 to 1 Reductions in volumes with sides 1/5 the cage width 100 to 1 or better in small volumes Whole Room cages may be provided Uniform field in center 34

35 Active Cancellation Cost If fluctuations in the ambient field can be kept below 5 mg at the site of the equipment, active cancellation provides sufficient attenuation for almost all charged beam equipment (SEMs, TEMs, FIBs). NYU is operating SEMs over LIRR tracks where fluctuations exceed 20 mg using active cancellation Equipment runs $30,000 to $40,000 per unit installed. Extremely reliable with practically zero maintenance. Many units have run nearly 20 years 24/7 without failure. 35

36 36 APPENDIX

37 The Basic Concept of the Hi-Lo Split Feeder Contact Wire Current H Track Current = Contact + Feeder Current D Feeder Current The Hi-Lo System works by creating a counter circulating current to cancel the field from the contact wire/track loop. Maximum attenuation is obtained attained when: D = H x feeder resistance contact wire resistance 37

38 Hi-Lo Feeders Concept, cont d. Significant Fields are only seen near the train where the current transfers between the feeders and the contact wire. And these fields are greatly attenuated. Contact Wire Current riser riser Feeder Current 38

39 Prediction of Distortion- Geomagnetic The predicted decay rate is applied to the distribution to generate a prediction of the distortion over a 2d plane The result is converted to a Mathcad matrix to allow the distortion at any point in a 2d plane to be evaluated. 39

40 Prediction of Distortion - Traction Current Field 40 Simplifications Assume all the current transfers from the feeder to the contact wire within 2 risers on either side of the pantograph Spice confirmed validity of the approximation. Discount the effect of the division of return current between the track and adjacent vehicle car bodies. Initial model accounted for this but we found the effect was negligible.

41 Prediction of Distortion- Combined The floor of the 3d graph is adjusted so that the distortion footprint may be seen for any specific magnetic field strength. Floor at 0.1mG Floor at 0.15mG Floor at 1mG 41

42 Prediction of Distortion- Combined The graph is tilted and rotated to get a top down view A contour of the distortion footprint at that field strength is obtained 0.1mG Contour 42

43 Prediction of Distortion- Combined The peak for limit values of distortion are plotted as contour lines along the alignment The effect of power system mitigation methods may be compared 0.1mG Contour without High- Low Mitigation 0.1mG Contour with High-Low Mitigation 43