Matiatia Marina Response to Councils Request for Further Information

Transcription

1 11 th May 2013 Waiheke Marinas Limited C/- Max Dunn Andrew Stewart By Dear Max, RE: Matiatia Marina Response to Councils Request for Further Information We refer to the Council s letter of 8 April requesting further information under Section 92 of the Resource Management Act (the Act) on the noise and vibration effects of the proposed Matiatia marina (Item 3 of letter). The Councils request was prepared with reference to our noise and vibration assessment dated the 6 th March 2013, submitted with the resource consent applications. It is important that our original report and the Councils s92 request are read in conjunction with this response for a complete understanding. The s92 requests are not repeated here but are responded to in the same alphabetical order: A) Construction Noise The s92 request seeks an explanation of why the Auckland City District Plan Hauraki Gulf Islands Section, Proposed Decision (2009) 1 (PDP) controls are applicable and / or appropriate, or alternatively an assessment of construction noise against the limits set out in NZS6803:1999 Acoustics Construction Noise. In terms of whether or not the PDP rules are applicable, we have demonstrated in our original report that the PDP controls are indeed the most appropriate for the proposal in the context of the very complicated set of rules that apply. Tables 4.1 and 4.2 of Rule of the PDP limit construction noise with reference to Short-term duration work or Typical duration work with the division of the two categories being dependent upon whether construction work at any one location will be up to 14 calendar days or greater than 14 calendar days respectively. The marina construction work falls into the typical duration work category according to the PDP 1 All appeals relating to acoustics have now been ruled on. As such, the text of this document should be given full weight regarding matters of noise. 1

2 rule. The construction noise from proposal complies with the Typical Duration part of the rule and would therefore be a Permitted Activity in terms of the PDP as set out in our original assessment. In a related regard, item L of the Council S92 request seeks clarification on the sound sources used to predict construction noise. As outlined later in our response to this matter, the last line of Table 2 of our original report allows for a hypothetical scenario whereby all sound sources would be operating simultaneously. However, with reference to the revised (and attached) Wardale Marine Industry Consulting Ltd (Wardale) construction programme, construction activities will for the most part be undertaken in series rather than in parallel. The original Wardale construction program in Figure 34 of the AEE also could have been interpreted to indicate that construction works will endure for a full 21 month period, which is not really the case. The revised program shows more clearly that there will in fact be a period of approximately 30 weeks of construction works, followed by a break of approximately the same period again, then a second period of construction enduring for approximately weeks. In our view, the very long break in the middle of the job (when there will be no noise) is sufficient to enable the project to be broken into two for the purpose of assessment against the construction noise limits and duration adjustments. This brings the overall duration of each part of the construction phase closer to the 20 week Typical Duration of works specified in NZS6803:1999. Given the very intermittent nature of works even when construction is at full pace, it is our view that the Typical Duration limits in the PDP rules remain appropriate. In any case, it requires the operation of all items of equipment noted in table 2 of our original report to generate a noise level higher than the Long Term (>20 weeks) noise limit specified in NZS6803:1999 of L Aeq 70dB. As set out in the revised Wardale program, it is envisaged that the construction sequence will require operations generally in series rather than all in parallel, thereby avoiding the possibility of all plant items operating simultaneously. On this basis, it is very likely that the limits for Long Term duration work as set out in NZS6803:1999 will also be complied with. B) Ambient Noise Measurements The information requested to expand on the data provided in Appendix 2 of our earlier report is tabled below: Logarithmic Mean Arithmetic Mean Day time ( ) L A10 48dB L A10 46dB L A95 37dB Night time ( ) L A10 42dB L Aeq 40dB L A10 37dB L Aeq 28dB 2

3 C) Noise from Vehicle Movements during Marina Operation In order to inform the prediction of movements to the car park, survey data from the Whitianga marina has been obtained. This marina is considered comparable to the Matiatia proposal in terms of likely vehicle movements. Car park activity data for this marina is attached as Appendix 1. We have treated day time movements using averages and night time movements using maximum values as there is provision for averaging of noise levels during the day time, but not at night. It should be noted that the Whitianga Marina caters for a number of commercial fishing boats that the Matiatia marina will not. Although we have not allowed for it, it is possible that the early morning vehicle movements at Whitianga over-represent the movements at Matiatia, as the commercial vessels leave before dawn much of the time. Night Time The peak hour for night time traffic is between 6am and 7am, with a daily average of 2.5 vehicles in the hour. It is possible that these could all occur within the last 15 minutes of the hour, (i.e. between 0645 and 0700), creating the worst-case scenario in terms of noise levels. Given the very small size of the car park, the noise from the vehicles themselves would not be present long enough to control the L A10 noise level at receiver 5. Notwithstanding, should two cars be left idling, (and using a conservative estimate of the sound power level / sound pressure level of a car idling derived from measurement of L AW 68 db / L A10 40 db at 10m) the noise level at receiver 5 would be less than 25dB. The location of the receiver 5 property (Aandewiel & King) is shown in our attached Appendix 4 of our original report. Based on our measurements of L Amax 10m from a car door slam, the L Amax at receiver 5 would be no greater than 50dB; easily compliant with the night time L Amax noise limit. The noise from people will be much less than this. We note that in terms of effects, the peak night time traffic periods occur after the commencement of ferry activity in the morning, and before the ferries cease at night. Day Time The daily average of the day time movements is 159 over the period from 0700 to 2200, equating to an average of 11 per hour. Based on the same rational as above, and If a peak 15 min value of 11 is assumed the noise level at receiver 5 will be less than L A10 50dB; readily compliant with the day time noise limit. 3

4 We note that the very small size of the car park, along with the separation distance to the nearest receivers and the screening afforded by the existing topography means that noise levels will be inherently low. Car parks of a larger size and hosting greater vehicle movements are able to comply with the same noise limits in other situations where the receivers are only in the order of 5-10 metres away. Although noise from the car park will at times be audible at the nearby receivers, we anticipate that the noise levels will be well below the noise limits in all cases, and will not contribute with other marina noise emissions to the extent that cumulative effects will be an issue either. D) Pine Harbour Measurement Standard The measurements of vessels at Pine Harbour marina in Section 5.2 of earlier report were undertaken in accordance with NZS6801:1991 Measurement of Sound. E) Pine Harbour Measurement Results To preface this response, the measured L A10 levels requested are in this case irrelevant to the prediction of noise levels, as they are of vessel pass-bys and not of vessels motoring slowly or idling while berthed. Accordingly, our predictions are based on the measured L Aeq of the passby immediately around the closest point of approach and the result has been treated as if it were a constant source in the proposed marina, (i.e. not varying in distance with respect to the receiver). The vessels at Pine Harbour passed the measurement point as they traversed a small arc around the microphone; thus they were at a relatively constant distance from the measurement point for sufficiently long to obtain a representative sample. From our measurements, the difference between the L A10 level and the L Aeq level for a boat that is idling or underway slowly is 1dB. Therefore, calculations made using the L Aeq levels in the table below will yield noise levels 1dB lower than the L A10 levels presented in our original report. The very small margin between the L Aeq and the L Amax levels supports this also. Overall, it is our view that this approach is conservative and that the same vessels operating in the marina would be likely to generate noise levels between 1-5dB lower than those we have predicted. The table below shows the measured L Aeq and L Amax levels of the vessels measured at Pine Harbour. 4

5 Measurement Sample L Aeq (db) L Amax (db) Direction & No. Of Boats Vessel Type Out + 1 In 2 Yachts (under power) Out 1 Launch Out 1 Launch Out 1 Trailer Boat Out 1 Launch Out 4 Trailer Boat Out + 1 In 2 Launches Out 3 Launch, trailer boat & yacht (under power) Out 1 Trailer Boat Out 1 Launch Out 1 Launch Out 1 Launch Out 1 Launch In 1 Yacht (under power) Out 1 Launch Lowest Highest Range 8 8 F) Range of Noise Levels Expected from Vessels in Matiatia Marina As a preface to this response, the vessels measured at Pine Harbour represent the typical range of vessels that will be expected to use the Matiatia marina. The level of noise generated by a vessel is not related to its size or type; it is instead dependent primarily on the exhaust design and the level of maintenance undertaken by the owner. Because of the vast variety of boat designs and exhaust configurations it is not possible to test the entire range of vessels that may use the marina. Instead, the only practicable approach is to test a representative range of vessels from a similar and nearby marina for the purpose of predicting noise levels. This is the same approach taken for road traffic, whereby predictions are based on measured levels of other roads using typical traffic noise levels. It is not possible or practicable to manage road noise emissions to cater for the loudest vehicle likely to use the road. 5

6 Further Noise Measurements We have supplemented our original measurements with a further set of specific noise measurements targeting the noisiest likely boat and one of the quieter launches likely to use the marina. A series of noise measurements were undertaken on the morning of the 1 st May 2013 at two locations (simultaneously) inside Matiatia Bay to accurately quantify the level of noise that vessels in the marina will generate. All measurements were undertaken in accordance with NZS6801:1991 using Bruel & Kjaer 2250 precision integrating sound level meters, calibrated before and after the measurements. The measurement positions and vessel positions are shown in Appendix 2 to this advice. The positions were chosen to represent the worst case in terms of vessel position with respect to receivers, with clear line of sight between sources and receivers at all times. The measurements were conducted between approximately 5am and 5.30am before the first ferries arrived from Auckland, (and therefore avoiding contamination of the measurements). The weather during the measurements was cool and calm, with full cloud cover and a westerly breeze of approximately 1ms -1. The vessels chosen for the measurements represent the loudest and quietest launches likely to use the marina: Vessel foot overall fly-bridge sport fishing launch. 660HP (330HP Cummins 6 cylinder turbo diesel x 2). Rear-facing above-water exhausts with no watercooled mufflers (straight-through design). Vessel foot overall sedan top classic launch with single 160HP Hino 4 cylinder diesel with side-exhaust close to waterline. The measurements comprised the vessels being started and idling in the locations noted in Appendix 2. The measurements were all approximately 5-8 minutes long for each location, with the noise levels remaining very steady over the measurements. Vessel Position Measurement 1 (Receiver 5) Measurement 2 (Dennerly Wool Shed) L A10 db L Amax db L A10 db L Amax db

7 The results of the further measurements show that our previous predictions were a minimum of 2dB too high and that the loudest vessel likely to use the marina would generate noise levels no higher than the relevant noise limit of L A10 41dB (in respect of receiver 5). G) Night Time Vessel Movements & Noise Effects This response should be read in conjunction with (c) above. From the Whitianga marina car park data, and on the basis that every vessel leaving the marina generates (on average) 2 vehicle movements, there is only likely to be one vessel leaving the marina in any 15 minute period during the night (in the busiest hour, or in other words - one vessel per hour). It will not be practicable to control the duration between departures or arrivals of vessels. H) Night Time Noise Level (L A10 43dB) Sentence 3, Paragraph 4 of page 15 of our report states: The acoustic effects of a boat leaving the marina very early in the morning could be clearly audible, but still not at a level that could be considered intrusive. As noted in response (F), the loudest vessel is likely to generate a level no greater than L A10 41dB as a result of our most recent measurements. Thus, the assessment has been based on the slightly lower level, but nonetheless holds for a level of L A10 43dB as well. A level of L A10 41 db is in itself a low level of sound. From our measurements, the equivalent L Aeq sound level is typically 1 db lower. This compares favourably to a value of L Aeq 45 db for Outside bedrooms under the World Health Organisation s (WHO) publication, Guidelines for Community Noise. This in combination with the small number of expected events (please refer to point c) results in our view that the level of sound is reasonable. Notwithstanding, the noise created by ferries using the wharf typically starts from approximately 0545 and noise from the marina is expected to be largely inaudible at any of the proximate receivers when a ferry is anywhere inside the harbour. Additionally, the night time noise limit of L A10 45dB is designed to be reasonable for continuous exposure during the night for every night of the year, (for mechanical plant as an example). In this case, a noise level of L A10 41dB will be present for only a very short time of the night and will only be generated by the noisiest of vessels. Most vessels will be much quieter. 7

8 The background sound level has only partial relevance in the assessment of an intermittent noise (such as a vessel departing the marina) and the level of intrusion over the background sound must be considered in terms of its temporal nature. I) Noise from Electricity Transformer, Refuse/Recycling & Sewage Pumpout Facilities A series of noise measurements have been undertaken to determine the noise level likely to be generated by the proposed sewage pump, transformer and recycling facilities. All noise measurements have been conducted in accordance with NZS6801:1991. Sewage Pump The sewage pump at Orakei Marina was measured on the 3 rd May 2013 whilst in operation. The noise level during the operation of the pump is very steady, and from the measurements we have derived a sound power level of L WA 78dB. The pump is not a particularly directional noise source with levels in all directions being substantially the same. Based on these measurements, a minimum separation distance of 18m between the pump and the nearest receiver is required to enable compliance with the night time noise limit of L A10 45dB. Allowing for the cumulative effects of other sources, (such as boats) we recommend that the minimum separation distance be 55m, resulting in a level of L A10 35dB. In this case, the pump is proposed to be located at the western end of the existing fuel dock, approximately m from Receiver 5. When the holding tank is full, it will be pumped out by a truck. We recommend that such pump outs are undertaken only between the hours of 7am to 10pm. Electricity Transformer An assessment of noise from a similar sized electricity transformer was undertaken at Orakei Marina on the 3 rd May At less than 1 meter from the transformer, no noise was audible or measureable above an ambient level of approximately L A10 40dB. The marina was almost full of vessels at the time of the assessment and nearly all vessels in this marina draw electricity from shore power connections. Based on our assessment we see no need for any mitigation of the proposed transformer. Recycling Facilities 8

9 We have undertaken numerous measurements of the disposal of bottles into recycling receptacles for other projects and also at Westhaven Marina. The L Amax noise level is typically as high as L WA 114dB. Such a level requires a minimum separation distance of approximately 35m to enable compliance with the night time noise limit of L Amax 75dB; a distance which is readily achievable in this case. Notwithstanding that compliance can be readily achieved by separation distance alone, we recommend that the recycling facilities are locked so they may not be used between the hours of 10pm and 7am on all days. If the receptacles have a top-hinged lid, it should ideally open in a westerly direction, thus minimising the noise levels to the closest receivers. J) Control of Halyard Slap & Wind Driven Electricity Generators Halyard slap is easily controlled by using bungee cords stretched around the mast and halyards. Placing a limit on the operational noise emissions from wind-driven electricity generators is both impracticable and unenforceable. As stated in our original report, halyard slap and the use of wind-driven power generators will be controlled under the marina rules. K) Underwater Noise Effects The s92 request seeks an assessment of underwater noise effects arising from the construction phase of the development and whether any noise control measures are required to mitigate potential adverse effects. We note that the s92 includes dredging as a source requiring consideration. In our experience, dredging will not generate any appreciable underwater noise; it is the piling and high energy activities only that require consideration, which in this case will only be piling. Given that the piling is situated in a bay that is frequented by large and noisy commercial ferries, we consider it prudent to put the predictions of underwater piling noise into context, taking account of the underwater noise generated by other significant anthropogenic sources. Existing Underwater Noise Environment In order to quantify the contemporary underwater noise levels generated by existing anthropogenic sources in the bay, a series of underwater noise measurements were undertaken by Styles Group in conjunction with Auckland University. Appendix 3 to this response includes the full measurement report from Auckland University s Institute of Marine Science, (the AU Report). From section of the AU Report, the measured source levels for the Fullers Quickcat and Starflyte ferries was 157dB re 1m and 151 re 1m. The levels were recorded whilst the ferries were inside the harbour travelling at 12km/hr and 11km/hr respectively. The ferries generate significantly higher levels once outside the harbour, with source levels of up to 9

10 185 re 1m recorded for the Fullers Superflyte ferry 3. Whilst the ferries were inside the harbour and travelling slowly or docking, the noise level at the inner-harbour measurement point was 145dB re 1μPa. Ambient levels in the absence of any proximate ferry or commercial sea traffic was as low as 123dB re 1μPa. With the very frequent commuter ferry movements throughout the day, along with other commercial and recreational boat traffic also operating inside the bay, the typical underwater sound environment is already affected to a considerable degree by anthropogenic sounds. Piling Noise The level of underwater noise from piling activities varies considerably depending on a number of factors, most notably the depth of water, sea bed composition, pile type and driving force. Unfortunately, it was not possible to undertake underwater noise measurements of piling in Matiatia Bay for this project. Also, there is no published data available on underwater noise levels for the combination of pile type, driving force, seabed type and water depth that will be found on this project. Most published studies are focussed on the effects of driving large diameter piles in deeper water for the construction of offshore exploration or drilling rigs, wind turbines and other large infrastructure. However, published measurement data suggests that piling in shallow water will be capable of generating peak source levels of between 170dB re 1m to over 235dB re 1m, with the range depending on the variables noted above. We note that 149 of the 160 piles to be driven for this project will be polyethylene-encased spun concrete piles that have a relatively high internal damping factor. The remainder of the piles will be tubular steel but will also have a polyethylene casing which will increase the internal dampening. The increased damping will reduce the high frequency underwater sound generation as well as the overall underwater sound levels. We note that there has been a lot of piling already undertaken inside the harbour with the construction of the ferry terminal and wharves. Most of the piles comprising these structures are considerably larger than those proposed for the marina and would have required a considerably greater driving force. Effects on Marine Mammals The effect of underwater noise on marine mammals has been widely studied but there is little certainty on the threshold of actual effects. Accordingly, conservative limits are often adopted. The most commonly adopted limit for the avoidance of behavioural effects from impulsive noise 3 M. Pine, A. Jeffs, C. Radford (2013) The Torment Beneath Planning Quarterly, Journal of the New Zealand Institute of Planning No. 188 March

11 sources is 160dB rms re 1μPa 4,5. For pile driving this equates to a peak level of approximately 175dB re 1μPa. Draft Unitary Plan Rule (2.1)(2)(a) of the Draft Unitary Plan states the following: Underwater activity must not exceed a noise level of 200 db re 1μPa measured at 1m from the noise source Whilst not in effect, the limit does provide some guidance in the absence of any other local rules or New Zealand Standards. The limit does not state whether the limit is a peak value or rms; we have taken the limit to be a peak value. Additionally, the limit is stated as a measured value at 1m. The employment of mitigation at a distance of 1m or greater would not therefore reduce the level. In our view this is an oversight of the rule, and the limit should simply be stated as a source level. Compliance with the limit should be determined based on measurements that take account of any mitigation measures employed, such as the presence of the breakwater. Notwithstanding the ambiguities of the rule in its current form, we recommend that piling be undertaken in accordance with the limit it proposes. Mitigation If necessary, the underwater noise from piling activities can be mitigated by employment of the following management techniques: (i) Using a soft start technique at the commencement of each pile being driven where the first drop is very small, working up to full height drops over a duration of minutes. The soft start requirements should be determined with reference to the measured underwater noise level of piling, with longer soft start durations required if the piling noise is within 10dB of 200dB re 1m and shorter durations if the underwater piling noise is less. (ii) By ensuring that piling does not commence if marine mammals are seen within 300m of the piling barge. This distance equates to a received underwater noise level of 160dB rms re 1μPa based on the piling being compliant with a peak level of 200dB re 1m. 4 United States National Oceanic and Atmospheric Administration Interim Sound Threshold Guidance Available at: hed_whales/killer_whales/southern_resident_killer_whale/section_7_consultations/interim_sound_thresh old_guidance.html 5 D. Peterson, D. Jurevicius, M. Warpenius (2011) Assessing the impacts of underwater pile driving noise on marine mammals 11

12 (iii) The implementation of any other physical mitigation measures that may be necessary, for example a reduced drop height for piling hammer or the use of a dolly between the hammer and the pile. Depending on the frequency of the underwater noise generated by piling and the type of piling rig being used, the physical mitigation measures could vary considerably in effectiveness. Overall, it is our opinion that the noise from underwater piling activities can be successfully controlled to within the proposed peak underwater sound level of 200dB re 1m, as measured beyond any mitigation specifically employed to reduce the levels. If compliance with this level is achieved, and if piling does not continue where marine mammals are sighted within 300m of the piling barge, the potential adverse effects of piling on the marine environment will be acceptable, and the level of underwater noise will be reasonable in terms of s16 of the Act. L) Predicted Construction Noise Levels The bottom line of the noise level predictions in Table 2 of our original report is based on one of each of the identified items of equipment working at the same time. This is consistent with the Wardale construction methodology report. M) Predicted Vibration Levels The prediction of vibration levels for the construction works was not undertaken in our original assessment for two reasons: Firstly, (and most importantly) it is never possible to prepare any predictions of vibration through the ground with any reasonable degree of certainty without site-specific measurements of the activity in question. This is because the propagation of vibration through the ground is extremely complex, especially in the marine environment and especially where there are elevation differences between the sources and receivers. Whilst the nature of the ground in the immediate marine environment is reasonably well understood due to the investigations that have been undertaken for this project, the nature of the ground beneath the receivers and in between is not well known at all, (to the degree necessary to inform vibration predictions) and the acquisition of such knowledge is not necessary or reasonable to require for a project of this nature. We have extensive measurements of ground vibration arising from piling on other projects. We have found through this work that the prediction of vibration attenuation through the ground is fraught with difficulty and nearly impossible to perform with any accuracy in the absence of sitespecific attenuation relationships. Also the plant specifications, hammer weight and drop heights have not been confirmed yet, and likely will not be until construction starts. 12

13 Secondly, the nearest dwellings are a considerable distance away (in terms of vibration propagation) and the ferry wharf is considered likely to be capable of withstanding high levels of ground movement. This means that the expense and difficulty of obtaining the necessary information through trial piling is unnecessary and impracticable. As mentioned in our original report, if vibration from construction activities is of concern, a pre-condition survey of the ferry building could be carried out and vibration monitoring undertaken at the commencement of piling at any of the proximate receivers to ensure compliance. The requirements of the German Standard DIN :1999 Structural Vibration Effects of Vibration on Structures (in its entirety) would need to be met. We note that there are several effective methods of mitigating vibration from piling should it become an issue. These including changing the drop height and hammer weight combinations, (for driven piles) and if vibro-hammers are used, the vibration frequency can be changed to avoid resonance effects if these are evident. For these reasons it is not practicable or indeed possible at this time to prepare any predictions of vibration arising from piling in this case. We suggest that the potential effects are best dealt with by way of monitoring and/or conditions of consent, according with the most common approach adopted in New Zealand. If a condition is to be imposed, we recommend the following: 1) Vibration generated by the construction of the marina shall comply with the requirements of the German Standard DIN 4150: Structural Vibration Part 3 Effects of Vibration on Structures at any dwelling or sensitive structure not under the ownership or control of the consent holder. At least 30 days prior to construction commencing the consent holder shall provide a Vibration Monitoring Programme to the satisfaction of the Team Leader Consents & Compliance. The programme shall identify how and when the vibration monitoring on buildings and structures located on properties adjacent to the marina will be undertaken during the piling and other relevant stages of the project and shall be prepared in accordance with recognised best practice and the German Standard DIN4150: Structural Vibration Part 3 Effects of Vibration on Structures. The vibration monitoring shall be undertaken by a suitably qualified and experienced person. A written record of the monitoring shall be provided to the Team Leader Consents & Compliance in accordance with the procedures identified in the approved programme. Note: Vibration measurements require access to buildings and structures located on adjacent properties and are therefore only able to be performed if access is granted / available. Should 13

14 access to a monitoring location be refused or otherwise unavailable, the requirements of this condition do not need to be met in respect of that building or structure. N) Assessment of Noise Effects on Dennerly Property (168 Delamore Drive) The s92 request seeks an assessment of noise effects for the residential receiver... adjacent to the foreshore... at 168 Delamore Drive, (near Measurement 2, Appendix 2). We understand that the building adjacent to the foreshore (as shown in Appendix 2) is an accessory building (former woolshed) and not a dwelling. Also, in terms of likely residential receivers we understand that there is an identified building platform located approximately 50m from MHWS due north of the end of proposed pier B. The approximate notional boundary of the identified building platform is a similar distance away from the development as the notional boundary of receiver 3 (from Appendix 4 of our original report) for all sources, including construction and vessels. On this basis, it is our view that any dwelling sited on the building platform on the Dennerly property is likely to be affected by noise emissions from the construction and operational effects of the marina to a similar degree as the other proximate receivers, including receivers 2 and 5. The further noise measurements detailed in response (F) show that vessel noise at night will be compliant with the noise limit of L A10 45dB. I trust that this information is satisfactory. Please do not hesitate to contact me should you have any queries or require any further information. Kind regards, Jon Styles Director & Principal Styles Group 14

15 Appendix 1 Car Park Movement Data from Whitianga Marina (Response C) 15

16 Appendix 2 Map of Further Noise Measurements (Response F) Measurement 2 Measurement 1 (Receiver 5) Boat Position 2 Boat Position 1 16

17 Appendix 3 Underwater Noise Measurement Report (Response K) 17

18 May 2013 Underwater Sound Assessment of Matiatia Bay, Waiheke Island Conducted by the Leigh Marine Laboratory, Institute of Marine Science For Styles Group Acoustic & Vibration Consultants Prepared by Matthew K. Pine Institute of Marine Science University of Auckland PO Box 349 Warkworth, NEW ZEALAND Mobile: Ph: ext 83622/ matt.pine@auckland.ac.nz

19 May Introduction Waiheke Marinas Limited is proposing to construct a small marina within Matiatia Bay, Waiheke Island; in place of the current swing moorings. Matiatia Bay contains a moderate ferry terminal which receives large commuter ferries from 0600 to 2300 hrs, daily. Our work aimed to characterise the ambient underwater sound within Matiatia Bay; in particular the anthropogenic sound transmitting from commuter ferries. 2.1 Methods and Materials Recording sites for vessels The sound emission from the Auckland commuter ferries, the Quickcat and Starflyte was recorded on 1 of May 2013, with good weather and sea conditions (11 km h -1 winds, < 10 % cloud cover and 0.2 m swell). Recordings were obtained within Matiatia Bay (S E ) and the Waiheke Channel (S E ) Recording system Recordings of commuter ferries were collected from the field using a calibrated High Tech, Inc. HTI-96-MIN omnidirectional hydrophone (10 Hz to 60 khz flat response) connected to a watertight temporal recording unit. The recording apparatus (containing the hydrophone, battery, recorder (20 db gain, 16 bit, 48 khz sampling rate) was bolted to a steel stand anchored by an iron bar and submerged in 5 and 12 m of water in low mean water. The marker float was anchored m from the recording apparatus to remove any extraneous sound from the float or rope.

20 May Data analysis Sound intensity spectra plots for a random 10 sec section of each recording, when the vessel was closest to the hydrophone, were generated to compare anthropogenic sound sources with each other and background sound levels. Each recording (total 4 per ferry) was subdivided in 5 subsamples, each 10 sec long, and third octave band analysis was carried out to compare anthropogenic sound sources with each other and background sound levels. The sound intensity spectra and octave band analysis for each anthropogenic sound were done using MATLAB software. Recording data (both anthropogenic and ambient recordings) was high pass filtered at 100 Hz to reduce interference by wind and surface waves, which typically transmit frequencies around 50 Hz. Mean sound levels inside Matiatia Bay were calculated for every 15 min from 0830 to Each 10 sec anthropogenic sound and control recording was band pass filtered into four different frequency groups: , , , and Hz. These bands were selected as most acoustic energy resides in the lower frequencies. For each frequency band, the mean sound level was calculated to compare anthropogenic sound sources with each other and background sound levels Calculating source levels and detection distances Source levels (SL) were back-calculated using the received level (RL) and transmission loss (TL). Transmission loss is the reduction of sound intensity as it travels through the water, and is defined by: TL(r) = SL RL(r) (Eq. 1) where r is the distance between the receiving hydrophone and anthropogenic sound source. Transmission loss was calculated using a modified equation from (Richardson 1995) which incorporates both spherical and cylindrical spreading, a published attenuation coefficient for

21 May 2013 shallow water, a spreading constant for shallow water and transmission losses due to depth, and is given by: 15 log RO 10 log R/RO log d / do (Eq. 2) where RO is the range in kilometres before transition to cylindrical spreading, R is the range between the source and the receiver in kilometres, d is the depth at the receiver in metres and do is the depth at the sound source in metres. Detection distances were calculated using the sonar equation: SE = SNR DT Eq. 3 where SE is signal excess, SNR is the signal to sound ratio, and DT is the detection threshold (Clark et al. 2009). A detection distance was defined as the distance at which the sound emitted from a source is loud enough to be detected by a generic fish species whose detection threshold is equal or less than 6 db re 1 µpa above ambient background levels (control). This threshold is conservative and has been assumed for a generic fish species in previous studies (Radford et al. 2005). Due to natural fluctuations of ambient sounds and other features of sound, a signal may not be detected by an animal, even when the signal itself is louder than the ambient level (Clark et al. 2009). The difference between the ambient sound level and the level at which an animal can detect the signal is termed a detection threshold (DT). Signal excess (SE) is the relation between DT and ambient sound levels (without targeted anthropogenic sound source) (NL) and is defined as the 50 percent probability of detection (Clark et al. 2009). Thus, the modified equation which was used for the calculation of detection distance was: SE = SL NL TL 6 (Eq. 4) where SE equals zero. Detection distances where subject to three important assumptions: (1) the sound and detecting receiver are omnidirectional; (2) detection threshold is 6 db above ambient

22 May 2013 background sound and; (3) ambient background sound (without anthropogenic sound source) is db re 1 µpa (measured NL for Matiatia Bay on 1 May 2013). 3.1 Results Background intensity levels The background sound intensity inside Matiatia Bay and the Waiheke channel was db re 1 µpa and db re 1 µpa, respectively Spectral levels The sound from the Quickcat ferry as it approached the Matiatia ferry terminal was largely of low frequency (< 2 khz) and showed a sharp peak at 600, 700, 800, 900 and 1000 Hz at approximately 116, 121, 132 and 136 db re 1 µpa Hz -1, respectively (Fig. 1). Sound generated by the Starflyte ferry was also of low frequency, with nearly all acoustic energy residing in frequencies below 5 khz (Fig. 1). The vast majority of sound from the Starflyte was below 1200 Hz with peaks at 1000, 500 and 100 Hz at approximately 110, 117 and 120 db re 1 µpa Hz -1. Control recordings of background levels without any operating vessels showed the majority of sound was between 1 and 20 khz. These results were also seen within the octave analysis data (Fig. 2a). Once the Quickcat ferry reached the wharf and the engines were in idle, only low frequency sound below 800 Hz was distinguishable from the control recordings (Fig. 2b).

23 May Mean sound levels Figure 1. Spectral composition of two underway ferries within Matiatia Bay and background sound levels without any operating vessels present. Mean sound levels varied significantly between 0830 and 1200 hrs; ranging from 122 db re 1 µpa during periods where no ferries were present to 145 db re 1 µpa as either the Quickcat or Superflyte ferry approached and departed the ferry terminal (Fig 3a). With no operating vessels within the bay, the overall mean sound levels were approximately 13 db re 1 µpa greater than ambient levels recorded in the outer Hauraki Gulf (ambient level around Leigh coast has been measured at 109 db re 1 µpa). Mean sound levels and frequency composition changed with increased distance from the source (Fig 4). Once the Starflyte ferry was inside Matiatia Bay, the overall mean sound level for khz at the closed receiving hydrophone (111 m distance) was 144 ± 0.5 db re 1 µpa and 143 ± 0.6 db re 1 µpa at Hz, 133 ± 2.5 db re 1 µpa at Hz and 124 ± 1.6 db re 1 µpa at Hz (Fig 4). At the same time, the hydrophone in the Waiheke channel (approximately 648 m from the ferry) detected the ferry at an overall mean received level of 125 ± 0.6 db re 1 µpa at khz, and 116 ± 0.4 db re 1 µpa at Hz, 112 ± 0.4 db re 1 µpa at Hz and 122 ± 0.6 db re 1 µpa at Hz (Fig 4).

24 Figure 2. Octave band analysis: (A) two underway ferries within Matiatia Bay; (B) Sound from the Quickcat ferry as it is underway (approach) and at the wharf unloading passengers (idle). The control is the recorded sound without any operating vessels present. May 2013

25 Figure 3. Recorded mean sound levels at khz within Matiatia Bay: (A) overall mean sound levels from two underway ferries (including background levels) and control recording; (B) overall mean sound levels every 15 min from 0830 to 1200 hrs. The dotted line represents background sound levels in the outer Hauraki Gulf. May 2013

26 May 2013 Figure 4. Octave band analysis for Starflyte ferry, as recorded from within Matiatia Bay (hydrophone 1) and from the Waiheke channel (hydrophone 2). The ferry was approaching the ferry terminal and was inside Matiatia Bay Source levels and detection distances A 6 db above ambient sound level hearing threshold was assumed for a generic fish species and audible frequency ranges below 5 khz. Therefore, the sound emitted from any anthropogenic source would need to be at least 128 db re 1 µpa before being detected by a generic fish species residing in the inner Hauraki Gulf. Detection distances varied between the two ferries, depending on the source spectrum and transmission loss, rather than source level alone (Table 1, Fig. 5). The Quickcat and Starflyte ferries had estimated source levels of 157 and 151 db re 1 1 m, respectively. The sound

27 May 2013 emitted from the Quickcat and Starflyte ferries was estimated to travel 338 and 305 m, respectively, before becoming undetectable in the inner Hauraki Gulf (Table 1, Fig. 5a). If cylindrical spreading was assumed (given by 10 log (r)), where r is the distance from the source), the detection distance of the Quickcat and Starflyte ferry was 1000 and 350 m, respectively (Fig. 5b). However, cylindrical spreading has been shown to poorly estimate true propagation of anthropogenic sound. Table 1: Source levels (db re 1 1 m) and theoretical detection distances for the Quickcat and Starflyte ferry. Detection distances are for a generic fish species and are based on a hearing threshold of 6 db above background sound levels and an audible freqeuency range below 5 khz

28 Figure 5: Level of anthropogenic sound (including background levels) recorded from within Matiatia Bay and the Waiheke channel for khz: (A) from the Starflyte ferry (actual measurement) and as estimated from a conservatively modified equation (Eq. 2); (B) from the Starflyte ferry (actual measurement) and as estimated by cylindrical and spherical spreading. May 2013

29 May Conclusions This investigation was carried out during a weekday morning in mid autumn. The weather was fine and only the Quickcat and Starflyte ferries were operating throughout the morning. Detection distances and estimated propagation of sound stated in this report are subject to many limitations and should be interpreted very carefully for a number of reasons: (1) Accurate hearing thresholds and frequencies are not confirmed for many fishes which inhabit the Hauraki Gulf; (2) The propagation of sound will vary in time and space due to weather, seasons, tidal currents, depths, bottom composition and ferry frequency. (3) The equations used in this study are based on data obtained from the field under different conditions. Matiatia Bay has a relatively narrow entrance and is surrounded by reef. The sound field generated by reflections, attenuation by bottom sediments and scattering of the sound introduces a degree of uncertainty. (4) Source levels and detection distances were estimated based on the assumption that depth increased consistently and the seabed was largely composed of soft sediments (sand). Side scan sonar data is unavailable and therefore specific bottom types in Matiatia Bay are unknown. (5) Background sound levels vary in time and space due to changing levels of boating activity throughout the year and biological activity (spawning seasons). All stated mean sound levels and source levels stated in this report include both the anthropogenic and background sound. The data from the investigation showed anthropogenic sound generated from the Quickcat or Starflyte ferry was prevalent from at least dawn till noon. Mean sound levels within Matiatia Bay varied from approximately db re 1 µpa to 145 db re 1 µpa.

30 May 2013 The Quickcat ferry had an estimated source level of 157 db re 1 1 m while travelling approximately 12 km h -1. The estimated detection distance of sound generated by the Quickcat was 338 m. The Starflyte ferry had an estimated source level of 151 db re 1 1 m while travelling at approximately 11 km h -1. The estimated detection distance of sound generated by the Starflyte was 305 m. The majority of acoustic energy from both ferries was below 5 khz. Several peaks below 1 khz were seen from the Quickcat ferry at a maximum power level of 136 db re 1 µpa Hz -1. As the Starflyte approached the ferry terminal within Matiatia Bay, the mean sound level within the frequency band Hz was approximately 144 db re m. This low frequency sound within Hz was detected by a hydrophone within the Waiheke channel at a mean sound level of 116 db re m. References Clark, C.W., Ellison, W.T., Southall, B.L., Hatch, L., Van Parijs, S.M., Frankel, A. & Ponirakis, D. (2009) Acoustic masking in marine ecosystems: Intuitions, analysis, and implication. Marine Ecology Progress Series, 395, Radford, C.A., Jeffs, A.G., Tindle, C.T., Cole, R.G. & Montgomery, J.C. (2005) Bubbled waters: The noise generated by underwater breathing apparatus. Marine and Freshwater Behaviour and Physiology, 38, Richardson, W.J., Thomson, D.H. (1995) Marine Mammals and Noise. Gulf Professional Publishing, Ontario, Canada.