Solar power mirror arrays for radio astronomy Olaf Wucknitz, Alan - PowerPoint PPT Presentation

Solar power mirror arrays for radio astronomy Olaf Wucknitz, Alan Roy wucknitz@mpifr-bonn.mpg.de aroy@mpifr-bonn.mpg.de Scintillometry Conference, Bonn, 7th November 2019

Solar power mirror arrays for radio astronomy • Solar Power Mirror Arrays • Phased array feeds • Test case J¨ ulich • Simulations of test observations • Issues, plans • Prospects O. Wucknitz 2019 2/33

Gemasolar as SKA? (Alan Roy, Ivan Camara, Olaf Wucknitz, …)

Gemasolar Basics Solar field: 2650 heliostats, each 120 m 2 , total 304 750 m 2 , equivalent to 620 m diameter single dish Tower height: 140 m Heat-Transfer Fluid: Molten salts (sodium + potassium nitrate) Receiver inlet temp: 290 °C Receiver outlet temp: 565 °C Turbine capacity: 19.9 MW Construction cost: 230 M€ (5 M€ from EU FP5, 80 M€ loan EIB) Timeline: 2007 begin, 2011 online Electricity sales: 110 000 MWh/yr = 30 M€/yr Ownership: Torresol Energy, subsidiary of consortium: 60 % SENER Grupo de Ingeniería (private company, Spain) 40 % MASDAR (alternative energy company of Abu Dhabi)

The Solar power array problem • many mirrors, different delays • signal spread over larger area • cannot catch the signal with one big feed • PAF ⋆ sample focal area ⋆ re-align phases ⋆ scale with signal strength � optimal weights from speckle pattern • Need to test concept! O. Wucknitz 2019 6/33

Some large facilities [ https://en.wikipedia.org/wiki/Solar_power_tower ] 2.6 km 2 collecting areas Ivanpah [ https://solarpaces.nrel.gov/ ] 1.1 km 2 Ashalim 1.2 km 2 Crescent Dunes 0.3 km 2 Gemasolar 0.018 km 2 J¨ ulich ( > 2 × Effelsberg) O. Wucknitz 2019 7/33

Solar Tower J¨ ulich: 150-m equivalent O. Wucknitz 2019 8/33

Solar Tower J¨ ulich: Experimental platform O. Wucknitz 2019 9/33

Practical issues • heat (use dedicated tower?) • RFI • mirrors: do they reflect radio waves? ⋆ expectation: must be thicker than skin depth ⋆ ca. 0.5 – 2 µ m for 10 – 1 GHz ⋆ metal mirrors (Gemasolar) okay ⋆ J¨ ulich: 0.2 µ m ⋆ actually seems to work! • can we predict speckle pattern? • beamforming techniques O. Wucknitz 2019 10/33

Technical setup for test in J¨ ulich • tripole antennas (Uppsala, Onsala) • simple uncooled receivers, mostly OTS parts • Rubidium/GPS clock • DBBC3 for sampling (512 MHz bandwidth, max. 6 channels) • Mark 6 or dedicated server for recording • correlate with Effelsberg for calibration (VLBI) • record, correlate, beamform, analyse O. Wucknitz 2019 11/33

Representative mirror positions for J¨ ulich 300 250 200 north [m] 150 100 50 0 200 150 100 50 0 50 100 150 200 east [m] O. Wucknitz 2019 12/33

Parameters • field size ca. D = 300m • 2150 mirrors, each 3.2 × 2.5m 2 • distance to receiver ca. L = 150m (ca. 50 m height) • assume 1.5 GHz (20 cm) • ‘focus’ size ca. 10 m • approximate speckle size ⋆ 0.2 m size ⋆ 1 MHz in frequency ⋆ 10 sec in time ⋆ 0.04 deg on sky O. Wucknitz 2019 13/33

Speckle image for J¨ ulich experimental platform 3 2 1 z [m] 0 1 2 3 4 2 0 2 4 x [m] O. Wucknitz 2019 14/33

Speckle dynamic spectrum for single feed 1550 1540 1530 freq [MHz] 1520 1510 1500 0 100 200 300 400 500 600 700 time [sec] O. Wucknitz 2019 15/33

Instrument response • mirror r m , focus/PAF element f p • time t and frequency ν generally omitted • delays c τ pm = | r m − f p |− θ · r m • voltage response for signal E ( t ) ⋆ time domain V pm ( t ) = g pm E ( t − τ pm ) V pm = E g pm e 2 π i ντ pm ⋆ freq. domain • total voltage response V p = E B p B p = ∑ g pm e 2 π i ντ pm m O. Wucknitz 2019 16/33

Beamforming theory ∑ B p V p p • fit of field E per t , ν E fit = | B p | 2 ∑ p • power estimate from all t , ν 2 � � ∑ � ∑ B p V p � � � t ν p S fit = | B p | 2 � 2 � ∑ ∑ t ν p O. Wucknitz 2019 17/33

Beamforming result: one antenna element, one sample beam map FT of beam 50 3.0 20 400 40 2.5 10 200 2.0 30 y [arcmin] V [m] 0 0 1.5 20 1.0 200 10 10 0.5 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] no position from one sample O. Wucknitz 2019 18/33

Beamforming result: 5 antenna elements, one sample beam map FT of beam 50 20 1.4 400 40 1.2 10 200 1.0 30 y [arcmin] 0.8 V [m] 0 0 20 0.6 200 10 0.4 10 0.2 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 5 elements spread over ∼ 5m O. Wucknitz 2019 19/33

Beamforming result: 20 antenna elements, one sample beam map FT of beam 50 1.0 20 400 40 0.8 10 200 30 0.6 y [arcmin] V [m] 0 0 20 0.4 200 10 10 0.2 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 20 elements spread over ∼ 5m O. Wucknitz 2019 20/33

Beamforming result: 100 antenna elements, one sample beam map FT of beam 1.0 50 20 400 0.8 40 10 200 0.6 30 y [arcmin] V [m] 0 0 0.4 20 200 10 0.2 10 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 100 elements spread over ∼ 5m O. Wucknitz 2019 21/33

Beamforming result: one antenna element, 100 samples beam map FT of beam 1.0 50 20 400 0.8 40 10 200 0.6 30 y [arcmin] V [m] 0 0 0.4 20 200 10 0.2 10 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 10 times ( ∆t = 10sec), 10 frequencies ( ∆ ν = 1MHz) O. Wucknitz 2019 22/33

Beamforming result: one antenna element, 2500 samples beam map FT of beam 1.0 50 20 400 0.8 40 10 200 0.6 30 y [arcmin] V [m] 0 0 0.4 20 200 10 0.2 10 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 50 times ( ∆t = 10sec), 50 frequencies ( ∆ ν = 1MHz) O. Wucknitz 2019 23/33

Beamforming result: one antenna element, 10000 samples beam map FT of beam 1.0 50 20 400 0.8 40 10 200 0.6 30 y [arcmin] V [m] 0 0 0.4 20 200 10 0.2 10 400 20 0.0 0 20 10 0 10 20 400 200 0 200 400 x [arcmin] U [m] 100 times ( ∆t = 10sec), 100 frequencies ( ∆ ν = 1MHz) O. Wucknitz 2019 24/33

Alternative beamforming E fit ∝ ∑ • formal result B p V p p B p = ∑ g pm e 2 π i ντ pm m � � E fit ∝ ∑ g pm e − 2 π i ντ pm V p ∑ • reorder m p c τ pm = | r m − f p |− θ · r m e 2 π i ν θ · r m / c ∑ e − 2 π i ν | r m − f p | / c V p E fit ∝ ∑ • split delay m p � two-stage delay beamformer (optical/analog?) O. Wucknitz 2019 25/33

Summary • solar power array radio telescope may actually work • PAF is essential, big PAFs not trivial • J¨ ulich: tests in preparation • many practical issues to consider • multi-beaming provides huge field of view • true Square Kilometre Array within reach !? • advanced beamforming, only cross-corr?, polarisation • build dedicated optimised array? • synergy with interstellar scattering/scintillation see backup slides from Alan Roy (scintillometry 2016) O. Wucknitz 2019 26/33

Model Mirror Locations

Dynamic Spectrum: Amplitude 80 MHz (1400 MHz to 1480 MHz) 12 min

Dynamic Spectrum: Phase 80 MHz (1400 MHz to 1480 MHz) 12 min

Secondary Spectrum 10 μs 1.1 Hz

Ivanpah Solar Power Facility (USA) O. Wucknitz 2019 31/33

Cerro Dominador (Chile) [ https://cerrodominador.com/ ] O. Wucknitz 2019 32/33

Some Google maps links https://en.wikipedia.org/wiki/Solar_power_tower • J¨ ulich https://maps.google.de/maps?ll=50.915,6.387778&t=h&z=15 • Gemasolar https://maps.google.de/maps?ll=37.558,-5.329&t=h&z=15 • Crescent Dunes https://maps.google.de/maps?ll=38.233,-117.366&t=h&z=15 • Ivanpah https://maps.google.de/maps?ll=35.57,-115.47&t=h&z=13 • Cerro Dominador https://www.google.de/maps?ll=-22.771,-69.485&t=h&z=15 O. Wucknitz 2019 33/33