EPA and OEH response to Jacobs SKM cost effectiveness review

Comment 1

The initial investigation into the proposed Newcastle Local Air Quality Monitoring Network (NLAQMN) monitoring locations undertaken by PAE Holmes (20110 included the results of PM10 dispersion modelling of emissions from industry located in and around the port. The results of modelling suggest Fern Bay may often experience higher impacts than other localities in particular higher than in nearby Stockton. The report by PAE Holmes (2011) recommend a monitoring station be located in Fern Bay. No such recommendation was made for Stockton. The network as proposed includes a monitoring station in Stockton and not Fern Bay. As such further consideration as to why Fern Bay should be excluded from the NLAQMN is recommended. While Fern Bay has a relatively low population the proposed increase in residential development in and around Fern Bay will see a rise in population in the future.

Response

The recommendation of the PAE Holmes report was considered. However Fern Bay has a significantly smaller population in comparison to Stockton, which is also closer to the port and the industrial area. The Newcastle Community Consultative Committee on the Environment (NCCCE) was briefed on all proposed sites and endorsed all locations.

Comment 2

One point of justification for the NLAQMN is to allow industry to rationalise the cost of monitoring. In this regard it is recommended that the EPA provide a framework for industry outlining how to undertake this rationalisation and remove redundant air quality monitoring from their regulatory requirements. This is needed to demonstrate the NLAQMN being cost effective and value for money.

Response

The EPA will take this comment under advisement. The EPA currently projects altering monitoring requirements for industry on an individual basis to ensure that all required monitoring is continued.

Comment 3

The NLAQMN funding model requires all licensed industry located within the Newcastle LGA that emits air pollutants which are to be monitored by the NLAQMN to contribute funds to the network in proportion to their emissions. It is noted that the investigations into the location of monitoring stations, namely the review of meteorology and modelling by PAE Holmes (2011), focussed only on industry within the “Newcastle Inner City and Port Neighbourhood”. It is recommended that further investigation is undertaken to assess industry source apportionment to impacts at each of the proposed monitoring stations as a means of allocating funds. This should be undertaken for all industry located within the Newcastle LGA.

Response

Attempting to apportion the impacts at each proposed monitoring station to specific industries is a difficult task. It would greatly complicate the management of this network and have significant cost implications. The network is also being established to assess the cumulative impact of industry and enable comparison during a range of wind conditions.

The review of meteorology and modelling by PAE Holmes only focussed on the Newcastle inner city and port neighbourhood. However, the EPA was required to determine an area to which the Regulation and subsequently the funding levy could be applied. The Newcastle LGA and State Waters in the Port of Newcastle were chosen as the most representative area from which industrial emissions contribute to the local air shed of the Newcastle area.

Comment 4

With respect to shipping emissions, the approach used by the funding model is to allocate these to industry that have ‘shipping in bulk’ scheduled activity in their Environmental Protection Licence (EPL). A suggested alternative approach is that costs are attributed to shipping movements (including ship type). This latter approach is recommended, as it will more accurately reflect total emissions from ships serving the individual industries.

Response

The funding model uses the commercial shipping emissions from the Newcastle Port reported to the Air Pollutant Inventory (API) and divides those emissions between premises that have the activity ‘shipping in bulk’ on their EPL. The distribution to each premises is based on the ‘shipping in bulk’ activity scale on the EPL.

It would be much more difficult to obtain actual emissions of each shipping movement from each licensee as third party information from the Port of Newcastle Authority or Newcastle Port Corporation would be required. Additionally, for some licences, the premise only includes a portion of the berth, or the infrastructure on the berth, not the actual berth.

Incorporating shipping emissions in the funding model and apportioning them is important given the likely increase in shipping activities in the port.

Comment 5

It is recommended that the EPA provide justification for using Tapered Element Oscillating Microbalances (TEOMs) for PM10 and Beta Attenuation Mass monitors (BAMs) for PM2.5, considering both cost and consistency of measurement technology, compared with using a combination of either BAMs or TEOMs for both particle size ranges.

Response

OEH has recently moved to a standard configuration for particle measurements across the NSW air quality monitoring network. TEOMs are used for continuous PM10 measurements and BAMs for continuous PM2.5 measurements. This configuration is active in the government-funded network and the industry-funded Upper Hunter Air Quality Monitoring Network. The proposed configuration for the Newcastle industry-funded network ensures consistency of PM10 and PM2.5 measurements across OEH operated networks. OEH notes that both the TEOM and BAM techniques meet Australian Standards for PM10 and PM2.5 measurements and there is little difference in cost between the two techniques. Each of the techniques has its own benefits/drawbacks, for example loss of volatile particles during sampling and sampling response times. These have been considered in the choice of monitors.

Comment 6

The independent review of costs revealed that in some cases the EPA itemised cost of equipment was ‘High’ – for example the costs for the portable monitoring enclosures and particle monitors. It is recommended that the EPA provide additional information with respect to the cost of these items so as to resolve the differences in cost estimates.

Response

The costs provided for the particle monitors are as quoted to OEH by the instrument suppliers and include costs for some instrument consumables. TEOMs are only available in Australia through a sole supplier. There are a few different BAM instruments available, but the Thermo 5014i is the most advanced model with a smart inlet heater and superior communications. Thermo 5014i instruments are used throughout the OEH network. The costs for the portable buildings reflect costs associated with building a portable laboratory to OEH design standards. The portable laboratory is more than just a portable building. It includes additional insulation to maintain greater internal temperature stability;  substantial internal wiring, including communication guttering and multiple LAN and COM ports throughout; and a rooftop work platform and access point that conform to OEH WHS requirements.

Comment 7

The EPA operational costs include annualised replacement costs for the various analysers, assuming they are replaced every four years. It is recommended the EPA provide some benchmarking of analyser replacement with consideration of their other existing monitoring stations.

Response

It is industry-standard practice to replace continuously-operating monitors every 4-7 years depending on the instrument type, manufacturer and operational history. For planning purposes budgets were prepared on a 4-year replacement cycle to provide the most conservative cost estimates. In practice, OEH expects that instruments within the network will most likely be replaced every 6-7 years.

Comment 8

It is recommended that the EPA provide clarification as to if or how the monitoring officer costs at $55,222 relate to the $50,000 one-off data management costs.

Response

The $50,000 data management costs relate to the capital expense of establishing web-based reporting for the network. This includes map and table-based data delivery, searchable data downloads and back end infrastructure costs. OEH notes that this one-off capital expense of $50,000 applies to the network as a whole, and not per station.