Scenario-Based Assessment of Water Quality and Ecological Impacts of Pump Station Overflows in a Peri-Urban Estuary

Abstract

Wastewater overflows (WOs) are a growing concern for water quality and ecological health in urban estuaries. This study provides a robust water quality and ecological assessment of WOs from four pump stations discharging into the Waimea Estuary, Aotearoa, New Zealand. Using overflow scenario modelling, baseline and event-based water quality sampling, and whole effluent toxicity testing, we assessed the potential impacts under conservative (2 h) and worst-case (24 h) overflow durations. Results showed that, even under worst-case conditions, the estuary’s natural dilution capacity exceeded the median dilution required to meet the 95% ecological protection level. Ecotoxicity was site- and season-specific, with amphipods and mussels showing sensitivity at some sites, while algal assays indicated nutrient enrichment rather than toxicity. Impacts were spatially limited and unlikely to persist beyond one or two tidal cycles. The estuary’s tidal exchange and resilient biota further mitigated risks. This method provides a transferable framework for assessing intermittent discharges in other coastal systems, especially those with high ecological value and infrequent discharge events.

1. Introduction

Urban wastewater overflows (WOs) are a globally pervasive and increasingly urgent environmental issue. These intermittent discharges, often triggered by infrastructure failures, stormwater inflows, or power outages [1,2], release untreated or partially treated wastewater into receiving environments, including rivers, estuaries, and coastal waters. These discharges release a mixture of contaminants, including physical, chemical, and microbial to terrestrial and aquatic ecosystems [3,4,5,6]. The United Nations estimate that WOs contribute up to 10% of total wastewater discharges in urban catchments with reticulated systems [7].

Impacts of WOs have been widely documented in North America and Europe, where they are linked to algal blooms, hypoxia, shellfish contamination, and beach closures [8,9]. These blooms result from nutrient enrichment and reduced water clarity, often linked to high microalgal productivity associated with WOs [10]. Wastewater overflows also introduce pharmaceuticals and personal care products, affecting aquatic organisms such as zooplankton [11,12]. Downgraded microbiological classifications of shellfish growing areas associated with WOs are not only a public health concern but also affect the viability of shellfish farming businesses [13]. These impacts may worsen due to ageing wastewater infrastructure, population growth, and climate change [14,15].

Impact assessment frameworks typically focus on continuous or frequent discharges and rely on post-event monitoring or hydrodynamic modelling. Such methods often lack baseline data and fail to capture the short-term, site-specific ecological responses characteristic of estuarine systems [16]. This limits their usefulness for informing pollution reduction plans and protecting sensitive environments.

Infrequent or low-volume WOs—such as those from pump station (PS) failures—are particularly difficult to assess. Their impacts are often transient and spatially confined yet may still threaten vulnerable estuarine habitats. Despite this, few studies offer transferable methods to assess the scale, persistence, and ecological risk of such events.

To address these gaps, we applied a scenario-based water quality and ecological assessment framework to WOs from four PSs discharging into the Waimea, a peri-urban estuary in Aotearoa, New Zealand. This estuary supports ‘Threatened’, ‘At-Risk’, and ‘Valued’ species, and is recognised as an Outstanding Site of Special Wildlife Interest. The Waimea also receives treated wastewater and faces increasing urban pressures.

Our approach integrated the following:

• Scenario modelling of overflow volumes and durations (2 h and 24 h conditions);
• Baseline and event-based water quality monitoring;
• Whole effluent toxicity (WET) testing using multiple aquatic species;
• Risk characterisation using the Environment Institute of Australia and New Zealand (EIANZ) ecological impact framework [17] and Burgman’s risk assessment principles [18].

This study provides a scientifically defensible and ecologically meaningful risk framework transferable to coastal systems with limited historical data and similar regulatory requirements such as the New Zealand’s Resource Management Act (1991) [19].

2. Materials and Methods

2.1. Study Area

The Waimea is a peri-urban estuary at the southern end of Tasman Bay/Te Tai-o-Aorere, New Zealand. Covering approximately 3300 ha, it is one of the country’s largest enclosed estuaries [20]. It connects to Tasman Bay through two tidal channels on either side of Rabbit Island (Figure 1). Tides are semi-diurnal, ranging from 1.8 m (neaps) to 3.4 m (springs) [21], and the water residence time varies from 0.6 to 11.6 days [21].

Estuarine habitats include mobile sands, fine sand flats, seagrass (Zostera) beds, mudflats, pebbles and cobbles, high shore flats, glasswort (Sarcocornia) beds, native rushes and sedges, Spartina beds, and subtidal and river channels [20]. The estuary supports over 50 waterfowl, 112 invertebrate, and 41 fish species [22]. These ecological values make the Waimea an Outstanding Site of Special Wildlife Interest and a high conservation priority under the Nelson Resource Management Plan [22].

The Waimea Estuary receives freshwater discharges from the Waimea River and other smaller streams [20], as well as stormwater and wastewater discharges from the southern part of Nelson City (Stoke and Tāhunanui) and the suburb of Richmond in the Tasman District. These areas are served by the Nelson Regional Sewerage Business Unit (NRSBU). While stormwater discharges contribute to the contaminant loading to the Waimea, assessing the water quality and ecological impacts from these discharges was outside the scope of this study.

The wastewater system conveys a mixture of domestic and trade waste (mostly food and wood processing industries) through a rising main that follows the estuary margin, then across the tidal channel to the Bell Island Wastewater Treatment Plant (WWTP). The plant serves an equivalent population of 109,000–125,000 and treats ≈ 17,000 m³ per day (2016 data). Treated effluent is discharged on the outgoing tide during the first 3 h. The network also includes five PSs, four of which are evaluated in this study: Wakatu PS, Saxton Road PS, Songer Street PS, and Airport PS. Each PS has dry weather storage (with varying wet-well and backup pump capacity), an automated alarm system, and an overflow point. Wakatu PS, Airport PS, and Songer Street PS discharge directly to tidal waters while Saxton Road PS discharges via a short ditch. The number of WOs from these PSs has reduced since 1996, and notably since an upgrade of PS systems was completed in 2013. Over the 2023/24 period, there was a single overflow event from NRSBU PSs [23].

2.2. Overflow Scenario Formulation

Hypothetical overflow scenarios were developed using available data on pump capacity, storage, and failure response times. Two WO durations were modelled:

• Conservative case (2 h overflow).
• Worst-case (24 h overflow).

Two operational failure types were considered:

• Scenario 1 (Sc1): One pump operational―representing power failure or multiple breakdowns. For this scenario, flow rate capacities of single pumps versus the dry and peak wastewater flows and WO volumes were estimated.
• Scenario 2 (Sc2): No pumps operational―representing mechanical failure in the back-up generator during power outage or catastrophic event.

Overflow volumes were estimated for both dry weather and peak flow conditions, with consideration of the 2-h and the 24-h resolution times and the available storage capacity at each pump station wet well, then used to determine WO risk for each PS.

2.3. Wastewater Characterisation

2.3.1. Baseline and Overflow Sampling

The physico-chemical and ecotoxicological characteristics of the wastewater were monitored at the four PS sites during baseline (no-overflow) and overflow discharge conditions. To characterise baseline conditions, wastewater and seawater samples were collected on four occasions (two dry-weather and two wet-weather surveys) at each PS wet well and their respective outfall site. Sample collection dates and associated environmental conditions are described in Supplementary Material S1.

In each survey, one 24 h composite sample of wastewater was collected at each wet well using an ISCO™ auto-sampler (Isco Inc., Lincoln, NE, USA). The auto-sampler was programmed to collect and composite a standard sample volume (minimum of 200 mL) every 30 min. Upon collection, the composite samples (10 L food grade plastic container) were sealed and immediately transported to the Analytical Services Laboratory at Cawthron Institute for analysis.

2.3.2. Laboratory Testing

Upon arrival at the Cawthron Institute laboratory, the wastewater samples were mixed prior to being aliquoted into sub-samples for quantification of physico-chemical parameters and whole effluent toxicity (WET) tests. Samples for physico-chemical analyses were sent by overnight courier to Hill Laboratories (Hamilton) in appropriate coolboxes with freezer packs. The physico-chemical parameters tested were as follows: chemical oxygen demand (COD), 5-day carbonaceous biochemical oxygen demand (cBOD5), nutrients (total nitrogen, Total Kjeldahl nitrogen, total phosphorus), total petroleum hydrocarbons (TPH), total suspended solids (TSS), volatile suspended solids (VSS), total metals (arsenic, boron, cadmium, chromium, copper, lead, mercury, nickel, zinc), total sulphide, fluoride, turbidity (Turb), dissolved oxygen (DO), pH, salinity, electrical conductivity (Cond), and temperature (Temp). Microbiological parameters and emerging contaminants were not tested because Nelson City Council’s Trade Waste (TW) Bylaw No 214 does not prescribe limits for these parameters (see Section 2.5.1). Details of analytical methods are listed in Supplementary Material S2.

Three WET tests on the WO samples were conducted at Cawthron:

• Standard guide for conducting static toxicity tests with microalgae using the marine green micro-algae Dunaliella tertiolecta (96 h) (ASTM E1218-21) [24];
• Standard guide for conducting acute toxicity tests on test materials with fishes, macroinvertebrates, and amphibians, using the burrowing amphipod (Paracorophium excavatum) (96 h) (ASTM E729-23) [25];
• Standard guide for conducting saltwater bivalve mollusc embryo-larval development test, using the blue mussel Mytilus galloprovincialis (48 h) (ASTM E724-21) [26].

Briefly, the marine green micro-algae were exposed in a static, microplate system (n = 5) to a dilution series of the test solution during the exponential growth phase over 96 h under 200 µE/cm², at 18.1 ± 0.8 °C on a rotary shaker at 100 rpm. The growth of the algae exposed to the test solution was compared with the growth of the algae in the control. Algal density in each test chamber was determined with a particle counter (Multisizer 4 Coulter Counter, Beckman Coulter, Lane Cove West, Australia). A test solution was considered toxic when a statistically significant, dose-dependent inhibition of algae growth occurred.

Ten individuals of the burrowing amphipods were exposed in a static glass jar (n = 3) to a dilution series of the test solution over 96 h (in the dark at 17.5 ± 0.4 °C). Mortality from immobility after probing was recorded in each test solution and compared to control.

Embryos of the blue mussel were exposed in 10 mL glass vials to a dilution series of the test solution (in the dark at 17.5 ± 0.4 °C; n = 5). The survival of larvae was determined by assessing the number of normally developed D-larvae (D-yield) characterised under the microscope after 48 h. The number of abnormal D-larvae in a test solution provides an indication of the embryo toxicity. Survival (D-yield) at each concentration of test solution was compared to survival in the control to assess the ecotoxicological parameters.

Test organisms were concurrently exposed to a series of concentrations of a reference toxicant to ensure consistency of their response and their sensitivity. Copper (as sulphate salt) was the reference toxicant for the marine green microalgae, and zinc (as chloride salt) was the reference toxicant for the burrowing amphipod and the blue mussel larvae.

Concentrations of the test solutions and reference toxicants producing an effect (lethal concentrations) on 10% and 50% of the population of the test organisms (LC₁₀ and LC₅₀) with associated 95% confidence intervals were determined by fitting a non-linear regression to observed data using the R (R core team, 2021) package “drc” [27]. A one-way analysis of variance was carried out, after ensuring normal distribution of data and homoscedasticity, to detect significant effects of concentrations of the test solution (p < 0.05), followed by a Dunnett test to determine the no-observed effect and the lowest observed effect concentrations (NOEC and LOEC, respectively) (p < 0.05).

2.4. Receiving Environment Monitoring

Baseline and Overflow Surveys

Water quality monitoring was conducted under baseline and post-overflow conditions on three occasions at each PS outfall site. In these surveys, seawater samples were collected for physico-chemical testing as described in Section 2.3.2.

At the PS outfall sites, levels of DO, water temperature, salinity, pH, electrical conductivity, and turbidity were measured using a handheld YSI multiparameter probe. Surveys assessed both near-field effects and far-field conditions at popular recreational areas. Samples were collected at nearshore sites (<50 m, 100 m, and 200 m from the discharge point), estuary channel sites (<50 m, 100 m, 200 m from the overflow), and contact recreation sites (Monaco Boat Ramp and Parkers Cove, also sampled on the shoreline; Figure 1). Estuary channel sampling was conducted from a kayak.

2.5. Impact Assessment Approach

2.5.1. Water Quality Assessment

Concentrations of physico-chemical parameters in wastewater (sulphide, boron, fluoride, arsenic, cadmium, chromium, copper, lead, mercury, nickel, zinc, and pH) were compared with Nelson City Council trade waste (TW) limits. Physico-chemical concentrations in seawater samples were compared with ANZECC & ARMCANZ (2000) guideline values (95% level of protection (LoP)) [28]. The physico-chemical concentrations in seawater were also compared with these guidelines and discharge permit limits to determine wastewater dilutions required to achieve an acceptable level of ecological protection at each PS receiving environment site.

Principal components analysis (PCA) was applied to explore spatial variation in water quality across sites.

2.5.2. Ecological Assessment

Ecological impacts were assessed based on the EIANZ ecological impact framework [17] and Burgman’s risk matrix [18]. These frameworks consider the magnitude or consequence of a discharge activity. The EIANZ approach focuses on potential risks to ‘Threatened’ habitats and taxa, while the Burgman framework focuses on the likelihood of an impact occurring. The ecological risks to highly valued taxa or habitats present in the Waimea Estuary were assessed to identify outstanding, high-value and sensitive substrata/benthic habitats, and ‘Threatened’ (and ‘At-Risk’) marine taxa, with emphasis on the PS sites. As a first step, a list of macrofaunal invertebrate species was compiled based on the literature [22] and cross-checked against lists of threatened species produced by the Department of Conservation and the International Union for Conservation of Nature [29]. Taxa with ‘Protected’ status were cross-referenced with the New Zealand Wildlife Act (1953) [30] and the Marine Mammals Protection Act (1978) lists [31]. The macrofaunal invertebrate taxa records were also assessed against sensitive taxa based on ecological values and habitats using the EIANZ guidance (EIANZ, 2018), and the risk of adverse impacts to the listed taxa and habitats was also determined using EIANZ (2018) and Burgman (2005). Summary tables of risk level and definitions of spatial scale of impact, likelihood and consequence impact categories and their corresponding scores, and confidence levels are presented in Supplementary Material S3. Impact scores were based on magnitude × likelihood, using spatial scale, severity, and confidence levels.

3. Results

3.1. Overflow Volumes

Saxton Road PS was the only pump station calculated to overflow in Sc1 (>20,617 m³) (

Table 1. Estimated wastewater overflow volumes and overflow dilutions at the four pump stations during peak flow and dry weather flow conditions. NA: no data available for these samples.

	2 h Overflow	24 h Overflow				2 h Overflow		24 h Overflow
Pump Station	Estimated Volume of Overflow (m3)	Estimated Volume of Overflow (m3)	River Volume (m3/Tidal Cycle) (h) DoC (2009)	Tidal Compartment (Neaps; m3) (i) Heath (1976)	Total Volume (h + i = j; m3) (j)	Dilution3 Factor	Ratio (1:X)	Dilution Factor	Ratio (1:X)
Peak flow conditions
Saxton PS (Sc1)	1570	20,617	848,160	31,000,000	31,848,160	0.000049	20,291	0.000647	1545
Wakatu PS (Sc2)	0	906	848,160	31,000,000	31,848,160	NA	NA	0.000028	35,152
Saxton PS (Sc2)	2952	37,207	848,160	31,000,000	31,848,160	0.000093	10,788	0.001168	856
Songer PS (Sc2)	521	7227	848,160	31,000,000	31,848,160	0.000016	61,168	0.000227	4407
Airport PS (Sc2)	666.9	11,765	848,160	31,000,000	31,848,160	0.000021	47,754	0.000369	2707
Dry flow conditions
Wakatu PS (Sc2)	0	0	NA	NA	NA	NA	NA	NA	NA
Saxton PS (Sc2)	486	7614	848,160	31,000,000	31,848,160	0.000015	65,531	0.000239	4183
Songer PS (Sc2)	15.7	1168	848,160	31,000,000	31,848,160	0.0000005	2,027,785	0.000037	27,280
Airport PS (Sc2)	0	2698	848,160	31,000,000	31,848,160	NA	NA	0.000085	11,804

) and had the highest estimated overflow volumes across both scenarios, reaching 37,207 m³ in Sc2 (Table 1). In contrast, Wakatu and Songer PSs had the smallest estimated overflows (<1200 m³). Estimated discharge volumes ranged from 1167 m³ (Songer PS) to 7614 m³ (Saxton PS) during dry weather and from 906 m³ (Wakatu PS) to 37,207 m³ (Saxton PS) during peak flow conditions (Table 1). The overflow discharge results for Airport PS suggest there would be no discharge at all under the 2 h dry weather flow scenario. The lowest estimated overflow volumes were from Wakatu PS; this station was estimated to overflow only during peak flow conditions over a 24 h timeframe, discharging 906 m³ of wastewater. All PS dry flow scenarios were calculated to be <50% of the PS capacities.

3.2. Wastewater Quality

The physico-chemical characteristics of wastewater sampled at the four pump stations (PSs) during both baseline (no overflow) and overflow conditions were assessed against the trade waste (TW) limits.

Table 2. Concentrations of physico-chemical parameters in wastewater samples collected at pump stations during baseline and overflow conditions, typical concentrations reported in the literature, and limits prescribed in the Nelson City Council Trade Waste Bylaw.

	Baseline (No Overflow) Conditions		Overflow Conditions		Typical Concentrations Reported in the Literature		TW Limit
Parameter (Unit)	Min–Max	Median	Min–Max	Median	Min–Max (Median/Mean)	Reference
Total Kjeldahl nitrogen (g/m3)	-	-	0.01–62	50.5	45–60 (53)	[32]	150
Total phosphorus (g/m3)	-	-	2.1–8.2	5.85	3.3–13	[33]	50
Total sulphide (g/m3)	0.043–29	0.2	<0.01–0.36	0.025	10–47 (30)	[34]	1
Boron (g/m3)	<0.05–2	0.11	<0.03–0.12	0.067	50–100	[35]	25
Fluoride (g/m3)	<0.01–1.5	0.052	0.06–0.4	0.11	0.20–1.11	[36]	5
Arsenic (g/m3)	0.00083–0.05	0.005	<0.0011–0.018	0.0042	0.0015 (Max) 0.001 (Median)	[37]	1
Cadmium (g/m3)	<0.00005–0.02	0.0005	<0.00005–0.00059	0.00021	0.0004 (Max) 0.0002 (Median)	[37]	0.5
Chromium (g/m3)	0.002–0.19	0.02	<0.0005–0.026	0.0038	0.07 (Max) 0.015 (Median)	[37]	5
Copper (g/m3)	0.021–2.2	0.07	0.0159–0.099	0.043	0.11 (Max) 0.075 (Median)	[37]	5
Lead (g/m3)	0.00105–0.18	0.007	0.00131–0.0129	0.00245	0.015 (Max) 0.008 (Median)	[37]	5
Mercury (g/m3)	<0.00005–0.007	0.00045	-	<0.00008			0.05
Nickel (g/m3)	<0.0021–0.072	0.0083	<0.0005–0.03	0.007	0.09 (Max) 0.02 (Median)	[37]	5
Zinc (g/m3)	0.045–23	0.13	0.046–0.5	0.115	0.4 (Max) 0.2 (Median)	[37]	5
Total petroleum hydrocarbons (TPH) (mg/L)	-	-	12.7–39	25	1500–1800	[38]	30
Total suspended solids (TSS) (g/m3)	-	-	116–13,000	215	50–800 (300)	[33]	1000
pH	5.9–9	7.2	6.5–8.2	7.12			6–9
5-day biochemical oxygen demand (cBOD5) (gO2/m3)	-	-	100–690	285	150–450 (250)	[33]	1000

(-) no data.

summarises the minimum, maximum, and median values for each parameter. Overall, the results indicate that all parameters were within TW limits under both baseline and overflow conditions. These results indicate that, even during overflow events, the wastewater discharged complied with the TW limits set by the Council.

A comparison of baseflow and overflow wastewater quality showed that while many parameters were lower under overflow conditions, the degree of dilution varied by contaminant. This pattern reflects the different sources and behaviour of contaminants in the sewer network. Parameters such as conductivity, nutrients, and some metals were substantially reduced during overflows due to stormwater ingress and greater hydraulic dilution. In contrast, suspended solids and turbidity were sometimes elevated, likely reflecting stormwater entrainment of particulates and resuspension within the network. Similarly, variability in COD and BOD responses between baseline and overflow samples may relate to differences in the relative contributions of trade waste, domestic wastewater, and stormwater inputs during high-flow events. These results suggest that overflow quality is not simply a diluted version of baseflow wastewater but reflects a dynamic mixture influenced by rainfall, catchment inputs, and hydraulic conditions within the sewerage system.

3.3. Seawater Quality near Pump Stations and at Contact Recreation Sites

The analysis of seawater samples collected from contact recreation (CR) sites and PS discharge-receiving environments during both baseline and overflow conditions revealed several trends. Across all sites, background water quality was generally consistent during baseline conditions, except for salinity, which was notably lower at the Songer and Wakatu PS sites compared to Airport and Saxton PS sites (

Table 3. Summary statistics of physico-chemical parameters in seawater samples collected at contact recreation and pump station discharge-receiving environments during baseline and overflow conditions.

			Baseline (No-Overflow) Conditions											Overflow Conditions
CR/PS	Dist (m)	Statistic	Cond	Sal	pH	Temp	DO (%)	DO	Turb	VSS	TSS	TN	TP	Cond	Sal	pH	Temp	DO (%)	DO	Turb	VSS	TSS	TN	TP
Monaco	0	Min.	32,382	25.0	8.1	11.7	97.5	7.8	7.6	1.5	14.0	0.24	0.02	24,308	19.5	8.0	11.1	99.2	9.0	-	-	-	-	-
	0	Max.	39,938	32.9	8.2	17.1	109.4	9.6	41.0	7.0	54.0	0.44	0.05	35,044	30.9	8.2	13.3	111.6	10.2	-	-	-	-	-
	0	Median	37,545	30.7	8.1	15.8	100.4	8.5	11.4	3.5	19.0	0.32	0.03	29,676	25.2	8.1	12.2	105.4	9.6	-	-	-	-	-
Parkers Cove	0	Min.	37,822	31.0	7.9	12.0	96.2	7.8	4.8	3.0	11.0	0.30	0.02	27,114	20.1	8.1	11.1	101.6	8.5	-	-	-	-	-
	0	Max.	44,388	32.9	8.2	19.5	105.9	9.3	19.1	5.0	25.0	0.99	0.06	38,499	31.6	8.2	12.5	109.2	10.0	-	-	-	-	-
	0	Median	39,294	32.5	8.1	15.9	103.6	7.9	8.5	3.5	18.0	0.37	0.04	32,807	25.8	8.2	11.8	105.4	9.3	-	-	-	-
Airport PS	50	Min.	20,928	15.7	7.9	11.2	61.7	4.7	5.2	3.0	13.0	0.21	0.02	-	-	-	-	-	-	-	-	-	-	-
	50	Max.	42,755	31.2	8.1	19.7	98.6	8.9	280.0	37.0	410.0	0.58	0.32	-	-	-	-	-	-	-	-	-	-	-
	50	Median	37,544	30.5	8.0	16.6	97.5	8.3	48.0 *	5.5	63.0	0.33	0.06	-	-	-	-	-	-	-	-	-	-	-
	100	Min.	25,213	19.3	8.0	11.6	86.8	6.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Max.	45,362	31.8	8.2	21.6	99.0	8.8	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Median	38,405	31.2	8.1	16.5	94.9	7.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Min.	23,329	17.8	8.1	11.6	60.7	4.6	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Max.	44,338	32.7	8.2	19.2	105.3	9.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Median	38,639	31.3	8.1	16.8	98.3	8.2	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Saxton PS	50	Min.	31,863	27.4	8.0	11.8	89.5	7.6	3.1	1.5	10.0	0.23	0.02	25,519	20.0	8.0	10.3	74.4	6.8	5.9	1.5	6.0	1.10	0.05
	50	Max.	42,532	32.6	8.2	17.8	102.3	8.2	105.0	10.0	101.0	0.69	0.11	32,549	22.2	8.2	13.4	106.9	10.3	145.0	18.0	190.0	1.55	0.17
	50	Median	39,437	32.0	8.1	15.7	94.4	8.0	17.4	7.0	41.5	0.29	0.04	29,034	21.1	8.1	11.9	90.7	8.6	75.5 *	9.8	98.0	1.33	0.11
	100	Min.	33,715	28.5	8.1	12.2	93.3	7.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Max.	43,232	33.4	8.2	18.3	100.1	8.2	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Median	40,861	32.7	8.1	16.0	93.9	7.9	-	-	-	-	-	41,127	26.1	8.2	11.2	99.8	9.3	114.0 *	18.0	157.0	0.89	0.14
	200	Min.	24,757	22.2	8.1	12.2	91.7	7.1	-	-	-	-	-	10,113	7.5	6.9	13.1	73.0	7.2	8.8	5.0	9.0	0.87	0.06
	200	Max.	43,579	33.5	8.2	18.2	97.1	8.2	-	-	-	-	-	26,620	21.6	8.3	13.3	104.4	9.7	110.0	13.0	153.0	4.20	0.14
	200	Median	40,961	32.9	8.2	16.3	94.9	7.7	-	-	-	-	-	18,367	14.5	7.6	13.2	88.7	8.5	59.4 *	9.0	81.0	2.54	0.10
Songer PS	50	Min.	22,059	15.9	7.6	11.1	90.5	7.3	9.9	3.0	20.0	0.40	0.03	-	-	-	-	-	-	-	-	-	-	-
	50	Max.	42,305	31.7	8.1	18.3	102.6	9.4	35.0	6.0	56.0	2.80	0.06	-	-	-	-	-	-	-	-	-	-	-
	50	Median	34,807	26.7	8.0	16.2	93.1	7.9	27.5 *	5.0	38.0	0.80	0.05	31,894	28.9	8.0	9.8	86.9	8.2	24.0 *	15.0	42.0	2.60	0.05
	100	Min.	35,131	28.4	8.0	10.9	88.7	7.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Max.	41,409	31.6	8.2	17.5	97.1	8.8	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Median	36,475	29.9	8.1	15.8	94.8	7.6	-	-	-	-	-	30,447	26.3	8.1	10.5	104.1	9.8	54.0 *	10.0	88.0	0.87	0.07
	200	Min.	34,011	26.8	8.0	10.9	91.9	7.4	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Max.	42,669	32.2	8.2	18.1	98.9	9.0	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Median	36,073	29.8	8.1	16.0	95.4	8.0	-	-	-	-	-	29,561	25.9	8.1	10.6	104.6	9.7	65.0 *	8.0	94.0	1.61	0.10
Wakatu PS	50	Min.	13,706	12.7	7.4	8.8	34.5	2.8	14.7	4.0	24.0	0.40	0.04	-	-	-	-	-	-	-	-	-	-	-
	50	Max.	38,299	29.9	8.1	16.6	93.8	8.9	210.0	14.0	250.0	2.50	0.23	-	-	-	-	-	-	-	-	-	-	-
	50	Median	33,138	28.0	7.7	14.9	84.0	7.5	46.0 *	8.5	67.0	0.63	0.08	-	-	-	-	-	-	-	-	-	-	-
	100	Min.	22,682	17.9	8.0	10.7	83.4	7.2	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Max.	41,071	31.8	8.2	17.1	99.7	8.7	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	100	Median	34,008	27.9	8.0	15.2	91.8	7.9	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Min.	27,160	22.2	8.0	10.9	90.2	7.5	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Max.	41,634	32.2	8.1	17.1	96.2	8.3	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
	200	Median	34,231	28.0	8.1	15.3	92.7	8.0	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-

CR: contact recreation; PS: pump station; Dist: distance between mean high water mark and sampling site; Min.: minimum; Max.: maximum; Cond: electrical conductivity (µS/cm); Sal: salinity (ppt); Temp: temperature (°C); DO (%): percentage of dissolved oxygen; DO: dissolved oxygen (mg/L); Turb: turbidity (NTU); VSS: volatile suspended solids (g/m³); TSS: total suspended solids (g/m³); TN: total nitrogen (g/m³); TP: total phosphorus (g/m³); (-): no data. Median results considered in the compliance assessment highlighted in bold. Cells highlighted in yellow denote results that exceeded the applicable limits for turbidity—10 NTU (ANZECC & ARMCANZ 2000 upper limit), total nitrogen—0.3 (ANZECC & ARMCANZ 2000 lower limit), and total phosphorus—0.03 mg/L (ANZECC & ARMCANZ 2000 lower limit); (*)—turbidity concentration is likely to reduce condition, alter foraging strategies, and cause gill deformation in fish [39].

). Variation in salinity among sampling sites provides an indication of mixing conditions within the estuary. Lower salinities at Songer and Wakatu PS sites reflect stronger freshwater influence from the Waimea River and local streams in the southern part of Nelson City and reduced tidal exchange in these upper estuary locations. In contrast, the higher and more stable salinities observed at Saxton and Airport PS sites indicate greater marine influence and more effective tidal flushing. Following overflow events, salinity levels decreased at all sites, indicating dilution of wastewater effluents due to river discharges and stormwater inputs on the shoreline during wet weather. Turbidity often exceeded the ANZECC & ARMCANZ (2000) upper trigger value of 10 mg/L, especially near Airport, Songer, and Wakatu PSs during baseline conditions, and near Saxton and Songer PSs during overflow events [28]. These elevated turbidity levels suggest potential impacts on aquatic life, including reduced water clarity and stress on fish species. Nutrient concentrations, specifically total nitrogen and total phosphorus, occasionally exceeded guideline values at some sites under both baseline and overflow conditions (Table 3), but effects were spatially limited and short-lived due to tidal exchanges. Overall, the results suggest that while transient water quality degradation was observed following WO events, the estuary had capacity for rapid dilution and recovery.

Figure 2 shows the PCA results of physico-chemical parameters measured at PS discharge-receiving environments under baseline (no-overflow) conditions. The first two principal components (PC1 and PC2) together accounted for approximately 79% of the total variance in the dataset, indicating a strong explanatory power of the selected variables. PC1 was primarily influenced by salinity and electrical conductivity, both of which had strong negative loadings, suggesting that these parameters were key differentiators among sites. PC2 was largely shaped by pH and dissolved oxygen (both concentration and percent saturation), which also exhibited strong negative loadings. Among the sites, Saxton PS showed the least variability in physico-chemical conditions, indicating more stable water quality, while Wakatu PS exhibited the greatest variability, likely due to fluctuating freshwater inputs or site-specific hydrodynamics. The PCA results highlight distinct environmental gradients across the pump station sites and underscore the influence of salinity and oxygen-related parameters in shaping water quality patterns in the estuarine receiving environments.

3.4. Effluent Toxicity

Table 4. Summary of ecotoxicity results of the amphipod and blue mussel embryo-larval development (BMD) tests presented as lethal concentration for 10% and 50% mortality (LC₁₀ and LC₅₀), no observed effective concentration (NOEC), lowest observed effective concentration (LOEC), and threshold effect concentration (TEC). The toxicity values for November 2020 (wet weather) could not be estimated for both the blue mussel and amphipod assays as the acceptability criteria for the reference toxicant were not met.

Endpoint	Date	Dry/Wet Weather	Pump Station (Acute Toxicity)
			Airport PS		Songer PS		Saxton PS		Wakatu PS
			P. excavatum	M. galloprovincialis	P. excavatum	M. galloprovincialis	P. excavatum	M. galloprovincialis	P. excavatum	M. galloprovincialis
LC10 (%) (95% CI)	20 June	Dry	2.2 (1.2–3.3) 1	0.28 (0.2–0.4) 1	4.8 (3.6–6.1)1	0.24 (0.16–0.32) 1	5.3 (2.6–8.1) 1	0.75 (0.6–0.9) 1	4.4 (2.8–6) 1	1.4 (1.1–1.6) 1
	21 April	Wet	15.8 (3.9–27.8)	X 2	6 (4.0–8.0)	X	17 (3.2–30.7)	X	2.8 (1.4–4.3)	X
	22 May	Dry	1.0 (0.4–1.6)	2.7 (2.1–3.1)	3.5 (3.1–3.8)	3.6 (2.6–4.1)	13.2 (9.4–17.1)	3.0 (2.7–3.2)	1.4 (0.8–2.0)	3.7 (3.5–3.9)
LC50 (%) (95% CI)	20 June	Dry	6.4 (5.1–7.7) 1	0.87 (0.4–0.6)1	6.8 (5.2–8.3) 1	1.3 (1.2–1.5) 1	13.6 (10.8–16.4) 1	2.1 (1.9–2.2) 1	9.1 (7.6–10.5) 1	2.6 (2.5–2.8) 1
	21 April	Wet	33.6 (25.5–41.7)	X	12.4 (10.4–14.5)	X	22 (16–28)	X	8.2 (6.5–9.9)	X
	22 May	Dry	5.2 (3.9–6.4)	4.9 (4.6–5.2)	11.2 (9.1–13.3)	5.7 (5.4–6.0)	17.3 (11.8–22.8)	3.3 (2.9–3.8)	4.9 (3.9–5.9)	5.4 (4.8–5.9)
NOEC (%)	20 June	Dry	2.6	0.2	2.6	0.2	5.1	0.78	2.6	1.56
	21 April	Wet	12.5	X	6.25	X	12.5	X	6.25	X
	22 May	Dry	1.56	0.78	3.13	0.39	1.56	3.13	3.13	3.13
LOEC (%)	20 June	Dry	5.3	0.4	5.2	0.4	10.1	1.56	5.2	3.13
	21 April	Wet	25	X	12.5	X	25	X	6.25	X
	22 May	Dry	3.13	1.56	6.25	0.78	3.13	6.25	6.25	6.25
TEC (%)	20 June	Dry	3.7	0.28	3.7	0.28	7.2	1.1	3.7	2.2
	21 April	Wet	17.7	X	8.8	X	17.7	X	4.4	X
	22 May	Dry	2.2	1.1	4.4	0.55	2.2	4.4	4.4	4.4
No toxicity dilution (x)	20 June	Dry	27	358	27	358	14	91	27	45
	21 April	Wet	5.7	X	11.3	X	5.7	X	22.6	X
	22 May	Dry	45	90	22	182	45	22	22	22

¹ Effective concentrations (EC₁₀ and EC₅₀) were used in replacement of lethal concentrations for the June 2020 samples. ² X indicates that the toxicity parameters could not be estimated as the blue mussel embryo toxicity assays were not performed.

summarises results of WET tests conducted to evaluate the potential biological effects of wastewater discharges from the four PSs. The results indicated variable toxicity across sites and sampling events. Amphipod and mussel assays showed highest sensitivity at Airport and Songer PSs, particularly during dry weather. For example, the LOECs for amphipods at these sites were as low as 3.13%, indicating a relatively high level of toxicity in undiluted samples. In contrast, samples from Wakatu and Saxton PSs generally exhibited lower toxicity, with higher NOECs and lower toxicity dilution factors.

The algal growth inhibition test did not indicate toxicity in any of the samples. Instead, all samples stimulated algal growth, suggesting nutrient enrichment rather than toxic inhibition. This response is consistent with the elevated nutrient concentrations observed in the wastewater, reflecting the organic-rich nature of the effluent.

Seasonal variation was observed. Toxicity was generally lower during wet events, likely due to stormwater dilution. However, some tests conducted during wet conditions were invalidated due to poor gamete quality in the mussel assays, limiting interpretation for those events.

3.5. Benthic Ecology of Overflow Sites

Values assigned for taxa and habitat were occasionally very high to high (see Supplementary Material S3). Habitats critical to seabirds, freshwater fish, and marine fish taxa occur near the PS outfalls. However, many of the taxa listed are marine wanderers which visit the estuary occasionally, and only within restricted areas (e.g., marine mammals and sharks will often be restricted to deeper channel areas). The resident estuarine biota are generally resilient to fluctuating salinity, turbidity, and nutrient levels, and well-adapted to the estuary’s dynamic conditions. Therefore, short-duration WOs are unlikely to cause long-term ecological impacts.

3.6. Wastewater Dilution Ratios and Ecological Protection

To assess ecological risks posed by the WOs, dilution ratios were calculated for key metals and metalloids in the wastewater (cadmium, copper, zinc, chromium, lead, and mercury), comparing observed concentrations against ANZECC & ARMCANZ (2000) 95% level of protection (LoP) guidelines and coastal permit (CP) limits for the Bell Island WWTP discharge [28].

The results, summarised in

Table 5. Median and maximum dilutions required for metal/metalloid concentrations to meet the ANZECC & ARMCANZ [28] 95% level of protection (LoP) guideline and the coastal permit limits.

Metal	Cadmium (n = 140)	Copper (n = 140)	Zinc (n = 140)	Chromium (n = 140)	Lead (n = 140)	Mercury (n = 132)
Dilution ratio required for 95% limit of protection 2
Maximum	1:3.6	1:1692	1:1533	1:43.2	1:40.9	1:50.0
Median	1:1.1	1:47.7	1:10.0	1:7.5	1:13.6	1:2.5
n 3	46	140	139	117	84	73
ANZECC LoP 95%	0.0055	0.0013	0.015	0.0044	0.0044	0.0004
Dilution ratio required to meet the coastal permit (CP) limit
Maximum	NA	1:14.7	1:23.0	NA	1:1.8	1:6.7
Median	NA	1:1.4	1:12.0	NA	1:1.7	1:1.7
n	0	15	2	0	2	5
CP limit	0.06	0.15	1	0.5	0.1	0.003

n—total number of samples; ² 1-part wastewater to X-parts receiving water; ³ number of samples with metal/metalloid concentrations above the Bell Island WWTP median concentration in the effluent. Data from all contributors were combined to reflect a worst-case dilution scenario.

, show that the dilution ratios required to meet the 95% LoP guidelines varied widely among contaminants. While maximum dilution requirements for copper and zinc reached 1:1692 and 1:1533, respectively, the median dilution ratios were far lower (e.g., 1:48 for copper, 1:10 for zinc). Most other metals required dilution ratios < 1:14.

When compared to the CP limits, the required dilution ratios were even lower. For example, copper and zinc required median dilutions of only 1:1.4 and 1:12.0, respectively, to comply with CP thresholds. In several cases, the concentrations were already below the CP limits, requiring no additional dilution. This suggests that, under most conditions, the wastewater discharges from the PSs would not exceed regulatory thresholds after reasonable mixing with estuarine waters. The term ‘reasonable mixing’ is used in the New Zealand Resource Management Act, and it refers to a reasonable timeframe for wastewater contaminants to dilute and disperse to concentrations consistent with meeting the ANZECC & ARMCANZ (2000) 95% LoP guidelines or CP limits, where applicable.

Importantly, these dilution requirements were assessed against the backdrop of the Waimea Estuary’s natural assimilative capacity. The estuary exhibits high tidal exchange and short residence times (0.6–11.6 days), which facilitate dispersion and dilution of contaminants. While complete mixing has not been assumed in this study, even under worst-case overflow scenarios—such as a 24 h discharge from Saxton Road PS (Sc2)—the available dilution in the estuary was estimated to be at least 1:856 (Table 1), far exceeding the median dilution required to meet ecological protection standards.

3.7. Overflow Risks from Individual Pump Stations

The risk profiles of the four PSs assessed in this study—Wakatu, Saxton, Songer, and Airport—vary depending on overflow likelihood, discharge volumes, contaminant concentrations, and site-specific hydrodynamic conditions. This section summarises the findings from scenario modelling, water quality analysis, and ecological risk assessment to characterise the relative WO risk for each PS.

3.7.1. Wakatu PS—Low Risk

• Overflow only predicted under Sc2 during peak flow (906 m³ over 24 h).
• High available dilution (1:35,152) (Table 1) which far exceeds that required to meet ecological protection thresholds (1:48).
• Minimal contaminant loads; low metals and organics.
• Minor and temporary impacts (e.g., colour, odour, turbidity) expected in the high tide zone.
• Presence of mobile and tolerant species reduce risk of ecological risk.

3.7.2. Saxton PS—Highest Risk

• Highest predicted overflow: 37,207 m³ in Sc2 (24 h) (Table 1).
• Highest contaminant loads: cBOD5, COD, TSS, nutrients.
• Moderate dilution in worst-case scenario (1:856), but sufficient to meet ANZECC & ARMCANZ [28] 95% LoP guideline.
• Risks include:

  ○

  Temporary oxygen depletion and visual/olfactory impacts.

  ○

  Smothering of benthic habitats during low tide discharges.

3.7.3. Songer PS—Low Risk

• Overflow likely only under Sc2 (up to 7227 m³ during peak flow).
• High dilution available (1:4407 or higher).
• Elevated cBOD5 concentrations may cause short-term hypoxia in estuarine organisms, particularly if discharges occur at low tide.
• Temporary impacts limited to high tide zone.
• Fast recovery expected due to flushing and presence of resilient species.

3.7.4. Airport PS—Moderate Risk

• Large potential overflow in Sc2: 11,765 m³ (24 h peak flow).
• High dilution capacity (1:2707 in worst case).
• Occasional exceedance of ANZECC 95% protection guidelines for metals (copper, zinc).
• Most other contaminants within guidelines.
• Short-term turbidity, odour, and aesthetic impacts expected near high tide zone.

4. Discussion

4.1. Environmental Impacts and Estuary Resilience

This study assessed the water quality and ecological impacts of WOs from four PSs discharging to the Waimea Estuary. A summary of potential environmental impacts is presented in

Table 6. Scale, persistence, and likelihood of water quality and ecological impacts of wastewater overflows on the Waimea Estuary.

Possible Water Quality Impacts	Affected Discharge Locations	Spatial Scale of Impacts	Persistence/Duration of Impact	Likelihood of Impact	Risk Level
Visual impacts
Increased turbidity in water column—reduces water clarity.	All: potentially higher at Saxton PS and Airport PS, due to large discharge volumes, and the higher suspended solids at Saxton PS.	Small	Temporary: unlikely to persist beyond one tidal cycle.	Likely	Low
Increased phosphorus and nitrogen—stimulates growth of algae and undesirable aquatic plants, micro-organisms and invertebrates (e.g., mosquitos).	All: particularly high at Saxton PS.	Small: limited to high tidal zone adjacent to outfall.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Likely	Tolerable
Unrefined pollutants/litter—visually unattractive.	All: potentially higher at Saxton PS and Airport PS, due to large discharge volumes.	Small: limited to high tidal zone adjacent to outfall.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Certain	Tolerable
Unpleasant odour
Increased organic matter/BOD concentrations—reduces dissolved oxygen levels as organics decay. Produces unpleasant olfactory properties.	All: cBOD5 particularly high at Saxton PS and Songer PS.	Small: limited to high tidal zone adjacent to outfall.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Certain	Tolerable
Toxicity/disturbance to estuarine flora and fauna
Increased suspended solids—deposited sediment affects flora and fauna habitats (smothering).	All: particularly high at Saxton PS.	Small: limited to high tidal zone adjacent to outfall.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Likely	Tolerable
Increased turbidity in water column—reduces water clarity, resulting in impact on fish and aquatic plants.	All: potentially higher at Saxton PS and Airport PS, due to large discharge volumes.	Small: limited to high tidal zone adjacent to outfall.	Temporary: unlikely to persist beyond cessation of discharge.	Likely	Negligible
Increased phosphorus and nitrogen—stimulates growth of algae and undesirable aquatic plants, micro-organisms and invertebrates (e.g., mosquitos).	All: particularly high at Saxton PS.	Small: limited to high tidal zone adjacent to outfall.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Likely	Tolerable
Increased ammonia, metals, COD—toxic to fish, benthic invertebrates, and aquatic insects at high levels.	All: COD particularly high at Saxton PS.	Small: limited to the edges of the overflow pipe.	Moderately persistent: around high tidal zone where tidal circulation is limited.	Likely	Tolerable
Cumulative impacts
Unrefined pollutants—contaminants—caused by combined action with other past, present, and future actions (e.g., accidental discharge + increased contaminated land runoff, storm water discharges, etc.).	All	Waimea Estuary	Moderately persistent: around high tidal zone where tidal circulation is limited.	Unlikely	Low
Simultaneous discharge—all four stations discharging at once.	All	From the Eastern Arm to the entire estuary	Persistent: around high tidal zone and below the low tide mark, where tidal circulation is more limited.	Unlikely	Low

. By integrating scenario analysis, field-based water quality monitoring, and ecotoxicological testing, the study offers a robust framework for evaluating the scale, persistence, and ecological risk of intermittent discharges in a dynamic estuarine environment.

The results indicate that the Waimea is resilient to short-term WOs under both conservative (2 h) and worst-case (24 h) overflow scenarios. The estuary’s natural assimilative capacity is sufficient to dilute contaminants to levels below ecological protection thresholds. Even in the most extreme case—Saxton PS discharging 37,207 m³ over 24 h—the available dilution (1:856) exceeded the median dilution required to meet the ANZECC & ARMCANZ [28] 95% level of protection (1:48). This finding is critical, as it demonstrates that the Waimea can buffer short-term discharges without breaching regulatory or ecological thresholds, provided that discharges are infrequent and not simultaneous across sites. Our results are broadly consistent with those obtained in other studies undertaken in peri-urban estuaries impacted by wastewater overflows. For example, a study in four estuarine systems in Greater Sydney (Australia) found no impacts of organic chemicals and trace metals tracked in wet weather ingress-diluted influent on ecosystem health, except for zinc and copper which were detected at concentrations above water quality guidelines, and ammonia which was detected at elevated concentrations in poorly flushed estuarine areas [40]. A study of trace metal inputs from combined sewer overflow discharges and metal mobility in an estuarine wetland in Queens (New York, NY, USA) found that metals bound to sediments can be remobilized and subsequently transported to downstream areas beyond the immediate discharge mixing zones while metals associated with plant roots may be effectively sequestered [41].

Other studies in temperate European estuaries found that nutrient pulses and associated oxygen demand effects from elevated discharge or runoff tend to be strongest in upper estuary or near-source zones, and that these impacts often attenuate with distance from the source via mixing, dilution, and sediment processes [42,43]). Although direct comparisons are limited, this suggests that in well-flushed estuarine systems—even under large inputs—contaminants such as nutrients are often transient and spatially constrained. The ecological risk assessment, based on the EIANZ [17] and Burgman frameworks [18], identified a ‘moderate’ risk to some high-value taxa and habitats, particularly in areas with limited tidal flushing or proximity to high-volume discharges. However, the overall likelihood of adverse ecological effects was assessed as ‘low’. This reflects the estuary’s high tidal exchange, short residence times (0.6–11.6 days), and the resilience of its biota, which are adapted to naturally variable salinity, turbidity, and nutrient regimes. The presence of mobile and hardy species, such as mud crabs, estuarine snails, and opportunistic macroalgae, further supports the conclusion that the estuary can recover from short-term disturbances.

Effluent toxicity testing revealed site-specific and seasonal variation in biological responses. Amphipod and mussel assays indicated higher toxicity at Airport and Songer PSs during dry weather, likely due to lower dilution and higher contaminant concentrations. However, these effects were largely mitigated by the estuary’s dilution capacity, and no persistent toxicity was observed beyond the immediate vicinity of the outfalls. The algal bioassays consistently showed growth stimulation, indicative of nutrient enrichment rather than chemical toxicity. This aligns with the elevated nitrogen and phosphorus concentrations observed in the wastewater and highlights the potential for short-term eutrophication effects, particularly in poorly flushed zones.

While microbiological contaminants were not directly assessed in this study, additional modelling by Hudson and Wadhwa [44] suggests that illness risks from PS overflows are generally low (<1%) across most of the estuary. Higher risks (>10%) were confined to drainage channels and depressions near discharge points, where tidal flushing is limited. This reinforces the importance of site-specific hydrodynamics in determining exposure and risk.

From a management perspective, the findings support a risk-based approach to overflow mitigation. While infrastructure upgrades may be warranted at higher-risk sites like Saxton PS, the overall low frequency and limited spatial extent of impacts suggest that targeted interventions—such as improved alarm systems, wet well capacity optimisation, and real-time monitoring—may be more cost-effective than large-scale capital works. Furthermore, the results provide a scientific basis for regulatory decision-making, particularly in relation to consent renewals and compliance assessments for wastewater infrastructure.

In summary, the Waimea Estuary demonstrates a high capacity to assimilate short-term, small-volume WOs without significant ecological degradation. However, this capacity is not unlimited. Continued monitoring, adaptive management, and proactive maintenance of PS infrastructure will be essential to ensure that the estuary’s ecological values are protected under future development and climate change scenarios.

4.2. Broader Application of the Assessment Framework

The integrated impact assessment method applied in this study—combining scenario analysis, baseline and event-based field sampling, and ecotoxicological testing—offers a robust and transferable framework for evaluating the environmental risks of wastewater overflows (WOs) in coastal and estuarine environments. This approach is particularly valuable in regions influenced by complex hydrodynamic and ecological conditions.

The method used in this study addresses these gaps in several important ways:

• Scenario-based risk framing: By modelling both conservative (2 h) and worst-case (24 h) overflow scenarios, the approach captures a realistic range of potential discharge events. This aligns with best practices in environmental risk analysis, where scenario testing is used to explore uncertainty and inform precautionary decision-making.
• Baseline and event-based sampling: The inclusion of pre-discharge (baseline) data allows for a more accurate attribution of observed changes to overflow events. This is a critical improvement for post-event-only studies, which may conflate natural variability with anthropogenic impacts.
• Ecotoxicological testing: The use of WET tests across multiple trophic levels (algae, amphipods, mussels) provides a biologically relevant measure of potential harm. This complements chemical analyses and helps identify sub-lethal or synergistic effects that may not be evident from contaminant concentrations alone.
• Dilution modelling and compliance assessment: The calculation of site-specific dilution ratios required to meet ecological protection thresholds (e.g., ANZECC 95% LoP) provides a quantitative basis for evaluating compliance and risk. This is particularly useful in estuarine systems where hydrodynamics vary significantly across space and time.
• Ecological contextualisation: The method incorporates habitat sensitivity and species vulnerability, using frameworks such as EIANZ and Burgman’s risk matrix. This ensures that assessments are ecologically meaningful and aligned with conservation priorities.

This framework is well-suited for application in other estuaries, lagoons, and semi-enclosed coastal systems, particularly those with high ecological or recreational value, intermittent or emergency overflow infrastructure, limited historical monitoring data, regulatory requirements for consent renewals or adaptive management.

5. Conclusions

By integrating scenario analysis, field-based water quality monitoring, and whole effluent toxicity testing, the study provides a robust framework for evaluating the scale, persistence, and ecological risk of intermittent discharges. The key conclusions that emerged from this study were

• Overflow risk varies by site: Saxton PS was identified as the highest-risk site due to its limited wet well capacity, high inflow volumes, and elevated contaminant concentrations. Airport PSs were considered moderate risk under the modelled scenarios while Wakatu and Songer PSs were assessed as low-risk, with overflows unlikely or well-mitigated by estuarine dilution.
• Dilution capacity exceeds requirements: Even under worst-case 24 h overflow scenarios, the estuary’s natural assimilative capacity (e.g., 1:856 dilution at Saxton PS) exceeded the median dilution required to meet the ANZECC & ARMCANZ [28] 95% level of protection (1:48). This suggests that the estuary can effectively buffer short-term discharges without breaching ecological thresholds.
• Ecological impacts are localised and temporary: Most predicted effects—such as increased turbidity, nutrient enrichment, and short-term oxygen depletion—were confined to the high tide zone adjacent to the outfalls and were unlikely to persist beyond one or two tidal cycles. The estuary’s resilient biota and high tidal exchange further reduce the likelihood of long-term ecological degradation.
• Toxicity is site- and season-specific: Effluent toxicity varied across sites and sampling events, with higher sensitivity observed in amphipod and mussel assays during dry weather. However, no persistent toxicity was detected beyond the immediate discharge zones, and algal assays indicated nutrient enrichment rather than chemical inhibition.
• Methodological strengths and broader relevance: The scenario-based approach, combined with baseline and event-based sampling, fills critical gaps in the literature by providing a more realistic and ecologically meaningful assessment of intermittent discharges. This method is transferable to other estuarine and coastal systems, particularly those with high conservation value or limited historical monitoring data.

Overall, the findings support a risk-based management approach that prioritises infrastructure upgrades at higher-risk sites while maintaining monitoring and preventive maintenance at lower-risk locations. The study also underscores the importance of integrating hydrodynamic, chemical, and biological data to inform regulatory decision-making and ensure the protection of sensitive coastal environments.