Investigating high frequency Acoustic Doppler Current ...

Defence Research and Development Canada External Literature (N) DRDC-RDDC-2020-N190 December 2020

CAN UNCLASSIFIED

CAN UNCLASSIFIED

Investigating high frequency Acoustic Doppler Current Profiler (ADCP) capabilities in measuring turbulence in tidal channels

Emma Shouldice DRDC – Atlantic Research Centre Len Zedel Memorial University of Newfoundland Angus Creech Heriot-Watt University International Conference on Underwater Acoustics September 9, 2020 Virtual Meeting Volume 40, 005001 (2020) Pages: 8 DOI: 10.1121/2.0001306 Date of Publication from Ext Publisher: September 2020

The body of this CAN UNCLASSIFIED document does not contain the required security banners according to DND security standards. However, it must be treated as CAN UNCLASSIFIED and protected appropriately based on the terms and conditions specified on the covering page.

Template in use: EO Publishing App for CR-EL Eng 2019-01-03-v1.dotm

© Her Majesty the Queen in Right of Canada (Department of National Defence), 2020

© Sa Majesté la Reine en droit du Canada (Ministère de la Défense nationale), 2020

CAN UNCLASSIFIED

CAN UNCLASSIFIED

IMPORTANT INFORMATIVE STATEMENTS

This document was reviewed for Controlled Goods by Defence Research and Development Canada using the Schedule to the Defence Production Act.

Disclaimer: This document is not published by the Editorial Office of Defence Research and Development Canada, an agency of the Department of National Defence of Canada but is to be catalogued in the Canadian Defence Information System (CANDIS), the national repository for Defence S&T documents. Her Majesty the Queen in Right of Canada (Department of National Defence) makes no representations or warranties, expressed or implied, of any kind whatsoever, and assumes no liability for the accuracy, reliability, completeness, currency or usefulness of any information, product, process or material included in this document. Nothing in this document should be interpreted as an endorsement for the specific use of any tool, technique or process examined in it. Any reliance on, or use of, any information, product, process or material included in this document is at the sole risk of the person so using it or relying on it. Canada does not assume any liability in respect of any damages or losses arising out of or in connection with the use of, or reliance on, any information, product, process or material included in this document.

1. INTRODUCTION

As concerns about greenhouse gas concentrations and the effects of climate change have grown, thedemand for renewable sources of power, that have lower environmental impacts, has increased. In streamtidal turbines are a viable and sustainable means of energy generation, that provide an alternative to non-renewable sources of energy. Tidal turbines have been successfully deployed and have been used to generateenergy in several locations globally including, La Rance, France (Merlin et al., 1982); Orkney, Scotland(Probert, 2011) and Shiwa Lake, South Korea (Cho et al., 2012).

Characterizing hydrodynamic properties in tidal channels is required for accurate estimates of site spe-cific energy generation potential. Site characterization is necessary when determining the site viability forin-stream turbine deployment (Gooch et al., 2009; Hay et al., 2015). Site characterization is also essentialfor estimating the effects of hydrodynamic flows on instrument longevity and performance, as accurate in-formation can contribute to optimisation of turbine design. Specifically, the turbine industry has a need forhigh accuracy measurements of turbulent velocities over long durations of time and with quantifiable errors.In-situ instruments such as shear probes and velocimeters although proven methods of collecting velocitymeasurements (Fer and Paskyabi, 2014; Lohrmann et al., 1994; Lueck et al., 1997; McMillan et al., 2016;Voulgaris and Trowbridge, 1996), are difficult to position in tidal channels due to large current speeds (uptoand in excess of 3 m/s). Using the concepts of Doppler shifted frequencies, ADCPs can be used to measureproperties of highly dynamic fluid flows (Lorke and Wuest, 2005; McMillan and Hay, 2017; Thomson et al.,2012). Bottom-mounted ADCPs are an effective alternative to in-situ instruments for measuring turbulencein the presence of high current speeds; allowing remote, unsupervised measurements over long periods oftime.

ADCPs have been used to collect measurements of turbulent velocities in tidal channels (McMillanand Hay, 2017; McMillan et al., 2016; Shcherbina et al., 2018; Thomson et al., 2012). McMillan et al.(2016) presents measurements of turbulence, using three different acoustic instruments in Grand PassageNova Scotia, a tidal site where efforts are on-going to implement in-stream tidal turbines. Grand Passageis approximately 4 km long, has a varying width between 800 m − 2 km and has a maximum depth of 30m (McMillan et al., 2016). Measurements were collected using an Acoustic Doppler Velocimeter (ADV), aNortek 1 MHz Signature AD2CP and a 600 kHz Teledyne RD Instruments Workhorse ADCP. McMillan etal. [2016] concludes that the proven ability of Doppler profilers to make remote turbulence measurementsin the presence of high current speeds can meet the needs of the tidal industry. However, a method forassessing the accuracy or limits of this data has not yet been developed. This study aims to consider thecapability of high frequency ADCPs in measuring vertical turbulence in tidal channels. A model of acousticbackscatter integrated with output from a computational fluid dynamics (CFD) model will be used to assessthe measurement quality in the presence of 2 m/s current speeds in an idealised rectilinear tidal channel.

2. ACOUSTIC MODEL

Models of sonar, sound reverberation, acoustic backscatter and sound propagation have been developedin both the field of oceanography (Crossley et al., 2017; Murray, 2014; Zedel, 2008, 2015) and medicine (Moand Cobbold, 1992). These models have been applied to questions of sonar design or configuration and usedto understand the performance limitations or instrument processes. The model of acoustic backscatter usedto simulate turbulence measurements in this study is based on the work presented by Zedel [2008,2015].Those studies simulate pulse-to-pulse coherent bistatic Doppler sonar configurations while confirming themodel’s ability to accurately statistically calculate acoustic backscatter.

The model of acoustic backscatter has a three dimensional spatial domain with adjustable boundaries.The model domain is populated with a fixed number of randomly positioned backscatter targets that will beused to calculate the acoustic backscatter as a function of time. The Doppler sonar backscatter, S(t), givenby Zedel (2015) is equal to the summation of individual backscatter contributions for each target:

S(t) =n∑i

ai s(t−

(rsi + rri

)C

), (1)

where, S(t) is the total calculated backscatter, n is the total number of backscatter targets, ai is the backscat-ter amplitude per ith target, s(t) represents the transmit pulse and rsi and rri are the distance from the ith

target to the source and to the receiver, respectfully.

Figure 1 is a model schematic that demonstrates how velocities are integrated into the time evolvingmotion of the acoustic targets (black dots). The instantaneous tidal velocity and time difference betweenacoustic pulse transmissions are used to update the target location (xi, yi, zi) by a factor of δxi, δyi, δzi. Thetargets acts as passive tracers.

x

z

x2, y

2, z

2

x1, y

1, z

1

x

z

y

Figure 1: Model schematic shows acoustic backscatter targets (black dots), overlaid on vertical velocitiesfrom the tidal channel data where shades of green (blue) are negative (positive) vertical velocities. Adiagram of acoustic beam directivity is shown in red.

The model has periodic boundary conditions to ensure that the number of backscatter targets remains con-stant as time evolves. Backscatter targets that are advected outside the domain boundaries are wrappedaround the axis and randomized in the other two dimensions which guarantees uniform spatial sampling ofthe domain (Zedel, 2015). The orientation and parameters of the acoustic transducer are customizable toallow for various geometries and configurations.

3. COMPUTATIONAL FLUID DYNAMICS MODEL

In previous work, the model presented in Shouldice (2019) used isotropic turbulence data from a directnumerical simulation (DNS) (Burns et al., 2018). The scaling required to implement this turbulent datasetdid not reflect the kinematic viscosity properties observed in tidal channels. The application of a CFD LargeEddy Simulation (LES) tidal channel data in this study allows for higher quality simulation results with a

wider range of meaningful analysis.

For a more effective representation of turbulence in tidal channels, velocity data was taken from a LargeEddy Simulation (LES) computational fluid dynamics model of an idealised tidal channel measuring 1 kmx 200 m x 30 m (Creech et al., 2017), shown in Fig. 2. The LES used the Synthetic Eddy Method (Jarrinet al., 2006) to generate inlet turbulence that matched ADCP measurements in a real channel, with a meanflow speed of 2m/s at 15 m above the channel bed. Profiles for mean velocity, Reynolds stress and eddylength scales were defined at the inlet to match turbulent flow in an energetic tidal channel. More details ofthese can be found in Creech et al. (2017). The The velocity field had 1 sample per second over a simulatedhour of flow and a spatial resolution of 1 m x 1 m x 0.33 m.

Figure 2: The Large Eddy Simulation of an idealised channel used for acoustic backscatter, coloured bythe velocity magnitude. Courtesy of Creech et al. (2017).

A subsection of the velocity data from the LES model was integrated into the acoustic model and usedto inform the time evolving motion of the backscatter targets described in Section 2 (see Fig. 1). For theanalysis, the velocity data was sampled across subsection of the model domain, measuring 10 m x 10 m x30 m. The velocity data was linearly interpolated to provide 4 Hz temporal resolution and vertical spatialresolution of 3 cm. It was determined through previous experimentation that high resolution velocity datawas required for accurate simulation using the acoustic model.

4. ACOUSTIC MODEL CONFIGURATION

The acoustic backscatter model was configured to resemble the Nortek AD2CP, as this instrument haspreviously been deployed to measure hydrodynamic properties in tidal channels, for instance McMillan et al.(2016) and Guerra and Thomson (2017). The high frequency and faster sample rate of the AD2CP has been

Table 1: AD2CP configuration parameters

Carrier Frequency 1 MHz Bandwidth 500 kHzTime between Pings 0.25 s Sample Rate 4 HzCollection Period 8 min Ambiguity Velocity 1.47 m/sPings per Ensemble 1 Averaged Range Bin Size 0.35 m

been demonstrated to perform better than instruments sampling at lower frequencies for measuring verticalturbulence in tidal channels (Shouldice, 2019).

The simulated AD2CP has five transducers, one vertical beam and four divergent beams at an angle of25o relative to the vertical axis (Nortek, 2018). Limiting the number of beams simulated decreases computa-tional cost. For this reason, only the vertical beam which is used for accurately measuring vertical velocitiesis simulated here. The total acoustic model domain is 10 m x 10 m x 7 m. The AD2CP is placed 3 mbelow the model domain measuring over a 7 m range the maximum vertical range from the transducer 10m (see Fig. 3), the backscatter contributions are a result of spreading, attenuation and beam pattern. Thisconfiguration is representative of a sea floor mounted upwards looking sonar collecting vertical velocitymeasurements at turbine hub height, approximately 7 m above the sea floor.

-150 -100 -50 0 50 100 150

Y Position (cm)

300

400

500

600

700

800

900

1000

Z P

ositio

n (

cm

)

Figure 3: Beam pattern for a vertical AD2CP beam as simulated by the acoustic model. Warm (cool)colours represent regions of high (low) backscatter contributions. Black dots shows the 3 dB width andblack x’s show the beams midpoints.

Sonar properties including carrier frequency, bandwidth and time between pings are customized to re-semble a broadband AD2CP configuration. Table 1 shows the main configuration parameters. The transmitpulse is broadband, the time delay between the pulses leading edges is 0.26 ms and the pulse has a chirpedfrequency 750 kHz to 1.2 MHz. The pulse length is 0.25 ms.

5. SIMULATION RESULTS

The AD2CP model simulation outputs backscatter amplitude, phase shift between the transmitted pulse pairsand pulse pair correlations. From the phase shift and the ambiguity velocity the velocity measurements canbe estimated (Zedel et al., 1996). The backscatter amplitude provides a measure of backscatter intensitywith range from the transducer. To assess the accuracy of the AD2CP in measuring vertical turbulence thesimulated result was compared to the expected measurement, or the reference turbulence. The referenceturbulence is the expected result known from the LES tidal model that was used to update the time evolving

position of the backscatter targets in our acoustic model.

The acoustic model has a collection period equivalent to 8 minutes of sonar operation. Figure 4 (a)shows the simulated measurement of vertical turbulence over a 2 minute interval, Fig. 4 (b) show the ex-pected velocity. Simulated data was averaged over 0.35 m in range and 2 pings to reduce sample noise.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (min)

6

7

8

9

Ran

ge

(m

)

-0.05

0

0.05

Ve

locity (

m/s

)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (min)

6

7

8

9

Rang

e (

m)

-0.05

0

0.05

Ve

locity (

m/s

)

a

b

Figure 4: Simulated AD2CP measurement of vertical turbulence between the 0 − 1 minute interval (a)compared to the reference turbulence over the same interval (b).

Performing a spectral analysis of the AD2CP vertical velocity measurements provides a method to eval-uate the turbulent characteristic resolved. The results can be considered in spectral space by applying theTaylor’s Frozen Field Hypothesis. Turbulent energy spectra were calculated for each range bin and areshown as individual spectra in Fig. 5 as the light pink dots. The 12 range bin spectra were averaged together(solid blue line in Fig. 5) to calculate an overall energy spectrum from the simulated vertical turbulent mea-surement. The same process was applied to the expected turbulence data to create a reference spectrum forcomparison (solid blue in 5).

The spectrum from the simulated measurement closely matches the reference spectrum at frequenciesbelow 0.28 Hz. The Kolmogorov spectrum predicts a −5/3 slope in the inertial subrange (dashed line in5), and there is good agreement between between the simulations and that prediction below 0.28 Hz. Atfrequencies above 0.28 Hz the spectrum flattens into a noise floor, this is an expected result as values abovethis frequency cut-off are not expected to be resolved by Doppler sonar measurements (Veron and Melville,1998).

6. CONCLUSION

In-stream tidal turbines provide a sustainable and viable way to generate renewable energy, howeverthe success of the tidal industry in part depends on our ability to accurately characterize the hydro-dynamicproperties of the site over long periods of time. Conventional instruments for measuring velocity such as

10-2

100

Frequency (Hz)

10-7

10-6

10-5

10-4

10-3

10-2

10-1

Sp

ectr

al D

en

sity (

m2s

-2H

z-1

)

Average Simulated AD2CP Measurement

Reference Turbulence from LES Model

Kolmogorov Slope

Simulated AD2CP Measurment per Range Bin

Figure 5: Turbulent energy spectrum for the AD2CP vertical beam compared to turbulence spectrumfrom the reference turbulence.

shear probes are difficult to position in tidal channels due to the large mean current speeds. An alternativemethod to collect velocity measurements is by using Doppler sonar systems that can remotely collect un-supervised data over long periods of time and are easier to position in the presence of high current speeds.This application of ADCP has been tested in the field but the accuracy limitations in these highly turbulentenvironments have not be quantified.

We have reported on a model of acoustic backscatter that uses output from a CFD model of a rectilineartidal channel to simulate the measurement of vertical turbulence and horizontal current speeds. A broadbandDoppler sonar system is configured to resemble the operational parameters of the 1 MHz Nortek AD2CP,which has been tested in the field (Guerra and Thomson, 2017; McMillan et al., 2016). Results are comparedspatially and spectrally to the expected turbulence known from the CFD tidal channel output.

The Nortek 1 MHz AD2CP was able to resolve broad turbulence structure with a vertically orientatedbeam. The system accurately recovered the -5/3 spectrum in the inertial subrange and the levels matchedthe input velocities supplied by the LES model demonstrating that accurate dissipation measurements arepossible. Further work will focus on comparative performance of the slanted beams and overall accuracy ofresolved horizontal velocities and the ability to extract extreme velocity events.

ACKNOWLEDGEMENTS

© Crown Copyright, 2020

REFERENCES

Burns, R., Eyink, G., Meneveau, C., Szalay, A., Zaki, T., Vishiac, E., Ravoori, R., Devincentis, B., Wu, Z.,Lemson, G., Chen, S., Lee, M., Malya, N., Moser, R. D., J, P., Livescu, D., and K, Y. P. (2018). JohnsHopkins Turbulence Database.

Cho, Y., Lee, J., and Jeong, W. (2012). The construction of a tidal power plant at Sihwa Lake, Korea. EnergySources, Part A: Recovery, Utilization and Environmental Effect, 34(14):1280–1287.

Creech, A., Borthwick, A., and Ingram, D. (2017). Effects of support structures in an LES actuator linemodel of a tidal turbine with contra-rotating rotors. Energies, 10(5).

Crossley, G., Alexandre, A., Parkinson, S., Day, A., Smith, H., and Ingram, D. (2017). Quantifying uncer-tainty in acoustic measurements of tidal flows using a ‘virtual’ Doppler current profiler. Ocean Engineer-ing, 137:404–416.

Fer, I. and Paskyabi, M. (2014). Autonomous ocean turbulence measurements using shear probes on amoored instrument. Journal of Atmospheric and Oceanic Technology, 31(12):474 – 490.

Gooch, S., J.Thomson, Polagye, B., and Meggitt, D. (2009). Site characterization for tidal power. IEEEOCEANS, 10:1–10.

Guerra, M. and Thomson, J. (2017). Turbulence measurements from five-beam acoustic Doppler currentprofilers. Journal of Atmospheric and Oceanic Technology, 34(6):1267–1284.

Hay, A. ., Zedel, L., Nylund, S., Craig, R., and Culina, J. (2015). The vectron: A pulse coherent acousticDoppler system for remote turbulence resolving velocity measurements. In IEEE/OEC Eleventh CurrentWaves and Turbulence Measurements Workshop. St. Petersburg, FL.

Jarrin, N., Benhamadouche, S., Laurence, D., and Prosser, R. (2006). A synthetic-eddy-method for generat-ing inflow conditions for large-eddy simulations. Int. J. Heat Fluid Flows, 27:585–593.

Lohrmann, A., Cabrera, R., Kraus, C., and ASCE, M. (1994). Acoustic-Doppler Velocimeter (ADV) forlaboratory use. Fundamentals and Advacements in Hydraulic Measurements and Experimentation, (351-365).

Lorke, A. and Wuest, A. (2005). Application of coherent ADCP for turbulence measurements in the bottomboundary layer. Journal of Atmospheric and Oceanic Technology, 22:1821 – 1828.

Lueck, R., Huang, D., Newman, D., and Box, J. (1997). Turbulence measurement with a moored instrument.Journal of Atmospheric and Oceanic Technology, 14(1):143–161.

McMillan, J. and Hay, A. (2017). Spectral and structure function estimates of turbulence dissipation ratesin a high-flow tidal channel using broadband adcps. Journal of Atmospheric and Oceanic Technology,34(1):5 – 20.

McMillan, J., Hay, A., Lueck, R., and Wolk, F. (2016). Rates of dissipation of turbulent kinetic energy in ahigh Reynolds number tidal channel. Journal of Atmospheric and Oceanic Technology, 33(4):817 –837.

Merlin, A., Sandrin, P., Gres, P., and Hilairet, M. (1982). Arga, the new operation model for the ‘La Rance’tidal power plant. Power Apparatus and Systems, IEEE Transcations on, PAS- 101(2):290–204.

Mo, L. and Cobbold, R. (1992). A unified approach to modeling the backscattered doppler ultrasound fromblood. Biomedical Engineering, IEEE Transactions, 39(5):450–61.

Murray, J. (2014). A theoretical model of linearly filtered reverberation for pulsed active sonar in shallowwater. Journal of Acousitcal Society of Amereica, 136(5):2523–2531.

Nortek (2018). Current profiler Signature 1000.

Probert, T. (2011). Orkney tidal power: A hotbed on the seabed. Power Engineering International, 19(1):18–22.

Shcherbina, A., D’Asaro, E., and Nylund, S. (2018). Observing fine-scale oceanic velociy structure with anautonomous Nortek acoustic Doppler current profiler. Journal of Atmospheric and Oceanic Technology,35(2):411–427.

Shouldice, E. (2019). Simulating turbulence measurements in high flow ocean regions with an acousticbackscatter model. Masters thesis, Memorial University of Newfoundland.

Thomson, J., Polagye, B., Durgesh, V., and Richmond, M. (2012). Measurements of turbulence at two tidalenergy sites in Puget Sound, WA. IEEE Journal of Oceanic Engineering, 37(3):363–373.

Veron, F. and Melville, W. (1998). Pulse-to-pulse coherent Doppler measurements of waves and turbulence.Jpurnal of Atmospheric and Oceani Technology, 16:1580–1598.

Voulgaris, G. and Trowbridge, J. (1996). Evaluation of the Acoustic Doppler Velocimeter (ADV) for turbu-lence measurments. Journal of Atmospheric and Oceanic Technology, 15(272-289).

Zedel, L. (2008). Modeling pulse-to-pulse coherent Doppler sonar. Journal of Atompsheric and OceanicTechnology, 25(10):1824 –1844.

Zedel, L. (2015). Modelling Doppler sonar backscatter. In 2015 IEEE/OES Eleveth Current, Waves andTurbulence Measurement (CWTM), pages 1–5.

Zedel, L., Hay, A., Cabrera, R., and Lohrmann, A. (1996). Performance of a single-beam pulse-to-pulsecoherent Doppler profiler. IEEE Journal of Oceanic Engineering, 21.(3):290–297.

DOCUMENT CONTROL DATA

*Security markings for the title, authors, abstract and keywords must be entered when the document is sensitive

1. ORIGINATOR (Name and address of the organization preparing the document. A DRDC Centre sponsoring a contractor's report, or tasking agency, is entered in Section 8.)

Acoustical Society of America Suite 300, 1305 Walt Whitman Road Melville, NY 11747-4300, USA

2a. SECURITY MARKING (Overall security marking of the document including special supplemental markings if applicable.)

CAN UNCLASSIFIED

2b. CONTROLLED GOODS

NON-CONTROLLED GOODS DMC A

3. TITLE (The document title and sub-title as indicated on the title page.)

Investigating high frequency Acoustic Doppler Current Profiler (ADCP) capabilities in measuring turbulence in tidal channels

4. AUTHORS (Last name, followed by initials – ranks, titles, etc., not to be used)

Shouldice, E.; Zedel, L; Creech, A.

5. DATE OF PUBLICATION (Month and year of publication of document.)

September 2020

6a. NO. OF PAGES

(Total pages, including Annexes, excluding DCD, covering and verso pages.)

8

6b. NO. OF REFS

(Total references cited.)

27

7. DOCUMENT CATEGORY (e.g., Scientific Report, Contract Report, Scientific Letter.)

External Literature (N)

8. SPONSORING CENTRE (The name and address of the department project office or laboratory sponsoring the research and development.)

DRDC – Atlantic Research Centre Defence Research and Development Canada 9 Grove Street P.O. Box 1012 Dartmouth, Nova Scotia B2Y 3Z7 Canada

9a. PROJECT OR GRANT NO. (If appropriate, the applicable research and development project or grant number under which the document was written. Please specify whether project or grant.)

9b. CONTRACT NO. (If appropriate, the applicable number under which the document was written.)

10a. DRDC PUBLICATION NUMBER (The official document number by which the document is identified by the originating activity. This number must be unique to this document.)

DRDC-RDDC-2020-N190

10b. OTHER DOCUMENT NO(s). (Any other numbers which may be assigned this document either by the originator or by the sponsor.)

DOI: 10.1121/2.0001306

11a. FUTURE DISTRIBUTION WITHIN CANADA (Approval for further dissemination of the document. Security classification must also be considered.)

Public release

11b. FUTURE DISTRIBUTION OUTSIDE CANADA (Approval for further dissemination of the document. Security classification must also be considered.)

12. KEYWORDS, DESCRIPTORS or IDENTIFIERS (Use semi-colon as a delimiter.)

Underwater Acoustics; Turbulence; 3D Model; Large Eddy Simulation

13. ABSTRACT/RÉSUMÉ (When available in the document, the French version of the abstract must be included here.)

Tidal channels are highly dynamic regions of the coastal ocean that exhibit strong turbulent behaviour. High resolution measurements of vertical turbulence and mean horizontal current speeds in these regions are required by the in-stream tidal turbine industry for site characterisation. Such data can be used by industry to improve estimates of tidal dissipation rates, energy generation potential and predictions of stress on critical hardware components. The in-situ oceanographic instruments that are used to make measurements of turbulence, such as shear-probes or Doppler velocimeters can be difficult to position in the presence of large horizontal current speeds (up to 3 m/s) that are characteristic of these active tidal channels. Sea floor mounted acoustic Doppler Current Profilers (ADCP) can remotely collect turbulence data for extended periods of time and provide an alternative to in-situ instruments. However, the accuracy and limitations of these ADCP turbulence measurements in the presence of high current speeds needs to be quantified. These limitations are explored using a model of acoustic backscatter integrated with output from a Large Eddy Simulation (LES) of an idealised tidal channel measuring 1 km x 200 m x 30 m. We simulate the direct acoustic measurement of the turbulent vertical velocities.