Sacchi et al. 2021 - Natural background levels of PTEs in Balangero asbestos-mine groundwater

Sacchi and colleagues derive natural background levels (NBLs) for potentially toxic elements (PTEs) in groundwater from the former Balangero chrysotile asbestos mine in serpentinised mantle peridotite of the Lanzo Massif, a “Contaminated Site of National Interest” 30 km northwest of Turin in the Piedmont region of North Italy. The study illustrates a modified Italian NBL workflow that uses all available measurements rather than per-station medians, in order to capture the high naturally occurring concentrations of Cr, Cr-VI, Ni, Mn and Fe that arise from water-rock interaction with ultramafic and metabasite host rocks, ore minerals, mine tailings and waste sludge. The work is environmental natural-background characterisation, not food, drinking-water-supply, or consumer-product occurrence evidence.

Key numbers

Sampling frame and hydrogeological setting

Four formations potentially host groundwater in the study area: the crystalline rocks of the Sesia-Lanzo zone (SL), the serpentinite of the Lanzo Massif (SERP), fluvioglacial deposits and their colluvial-eluvial covers (FGD), and the alluvial deposits of the Balangero Plain (BPA). The first two are fractured hard-rock formations with discontinuous permeability; the latter two are porous deposits, with the BPA constituting a phreatic aquifer along the Stura River valley floor downstream of the remediation site.

The complete hydrogeochemical characterisation required for NBL assessment was conducted in 2019 on 30 monitoring stations (MSs), with sampling performed in spring (25 June-4 July 2019) after a rainy period. One station (APE040, BPA) exceeded the nitrate pre-selection threshold of 37.5 mg/L (75% of the 50 mg/L drinking-water TV) and was eliminated, leaving 29 MSs for statistical treatment. Earlier sampling campaigns starting in 2012 contributed additional PTE measurements pooled in Table S3 of the supplementary materials.

Italian guideline TVs for the PTEs of concern are Cr = 50, Cr-VI = 5, Co = 50, Ni = 20, Mn = 50, Fe = 200 and Zn = 3000 µg/L.

Hydrochemical facies

Field parameters across the dataset span a circumneutral pH range of 6.17 to 8.70, Eh values between 191 and 444 mV, and dissolved oxygen contents from 1.90 to 10.0 mg/L. Most samples are classified as oxic; APP127 (FGD) and APF140 (SERP) are closer to the limit of reducing waters. The dominant hydrochemical facies is Mg-HCO3, evolving towards a sulphate-rich Mg-SO4 facies with increasing mineralisation. Two SERP samples (APF138 and APF139), collected at the foothills of the Fandaglia waste dump, are of Mg-SO4 facies; one SL spring (ASC077) is of Na-Cl facies and is the only sample interpreted as influenced by road de-icing salt. Sample APF082 (SL), collected close to a rodingite vein, is of Ca-HCO3 facies. Electrical conductivity and sulphate are significantly correlated (N = 30; R2 = 0.67; p < 0.01).

Representative PTE values per monitoring station

Table 1 reports the representative concentration (median of measurements where N > 1, single measurement otherwise) for each of the 29 retained MSs. Italian TV exceedances are flagged in bold in the source; this page summarises the descriptive statistics rather than transcribing all 29 rows.

Table 2 descriptive statistics on the per-MS representative values (in µg/L, ND values substituted with half the LOD of 0.5 µg/L for all PTEs except Cr-VI which uses 1.5 µg/L):

Across the per-MS dataset the proportion of non-detect (ND) values is 28% for Cr, 79% for Cr-VI, 97% for Co, 10% for Ni, 41% for Mn, 41% for Fe and 62% for Zn.

Table 4 descriptive statistics on the full pooled dataset (all available measurements across all campaigns, including ND, in µg/L):

The source itself flags Co as unreliable from this dataset: the paper’s own narrative on p.13 states “Concerning Co, the statistics are not relevant as this element shows only two measurements > LOD, both from the same MS.” Downstream synthesis should treat Co as not quantifiable from Table 4; the per-MS Co statistics in Table 2 (97% ND, max 2.9 µg/L) are the correct anchor for Co from this paper.

Proposed natural background levels (Table 9)

The authors propose using all available measurements (Table 7) rather than per-MS medians (Table 6), having excluded the two MSs with low redox potential (APP127 FGD and APF140 SERP) on geochemical grounds. The resulting NBL set is:

The 95p is selected for Cr, Ni, Mn (and Cr-VI), and the 90p is selected for Fe and Zn, following the 1-1/N probability-level rule in the Italian NBL guidelines for the available number of measurements. The Cr-VI NBL is explicitly flagged as not representative of the BPA formation because detectable Cr-VI is restricted to six MSs mostly located at the mountain foothills.

The dataset is classified as type B (significant spatial dimension, inadequate temporal dimension) under the Italian guideline framework, with N > 25 stations and surface extension > 700 m^2, giving a medium confidence level on all NBLs except Cr-VI.

Comparison with similar hydrogeochemical settings

Comparison Italian regional dataset (Table 8) of groundwater unaffected by anthropogenic contamination (nitrate < 37.5 mg/L) from five variably serpentinised ultramafic areas (Table reproduces published statistics):

The Balangero 95p values for Cr (39.3) and Ni (84) fall between the 95p and 99p of the Italian comparison dataset; Cr-VI 95p (38.1) is lower than the comparison 95p (42.75). Fe and Mn 95p concentrations are higher than the comparison 99p but lower than its maxima.

The paper also reproduces (Figure 4) a regression of Cr-VI against total Cr for over 918 California drinking-water datapoints (Cr-VI/Cr slope close to 1), the Anthemountas Basin Greece dataset (slope 0.9425, R2 = 0.996), and the Balangero study site (slope 0.7161, R2 = 0.7232), supporting the interpretation that most Cr in solution at Balangero is in the hexavalent form.

Trend analysis

The dataset is inadequate to detect temporal trends per the Italian guidelines (no MS has eight half-yearly measurements at a regular frequency over at least two years). Trend tests at APP126 (FGD, 11 measurements) and APE031 (BPA, 10 measurements with missing years) did not reject the trend hypothesis for Cr, Cr-VI, Ni or Mn at APP126; at APE031 a Sen’s slope estimator gave a decreasing Ni trend of -7.87 µg/L per year and an increasing Mn trend of +0.27 µg/L per year when 2013 and 2014 missing values are interpolated. The authors conclude the trends are not sufficiently sound to detect changes in time, and neither station is excluded from the spatial NBL evaluation.

Methods (brief)

PTE analyses (Cr, Cr-VI, Co, Ni, Mn, Fe, Zn) were performed at the Nuovi Servizi Ambientali S.r.l. laboratory (now Lifeanalytics Torino S.r.l.) using inductively coupled plasma optical emission spectrometry (ICP-OES) for all PTEs except Cr-VI, which was analysed by ultraviolet-visible spectroscopy (UV-VIS). Detection limits are 1 µg/L for all PTEs except Cr-VI (3 µg/L) and Al (1 µg/L); values < LOD were substituted with half the LOD. The paper does not name the ICP-OES or UV-VIS instrument vendor or model.

Field measurements (temperature, redox potential, electrical conductivity, pH, dissolved oxygen) used the HACH HQ40D multiparameter probe with a flow chamber to avoid air contact; field-measured Eh was corrected to the standard hydrogen electrode. Alkalinity was measured in the field with the HACH Digital Titrator and standardised titration cartridges. Anion samples were collected in high-density polyethylene (HDPE) pre-washed containers; cation and trace-element aliquots were field-filtered to 0.45 µm and acidified with ultrapure hydrochloric acid; hydrocarbon samples were collected in dark glass containers. Wells were purged 3-4 well volumes with a low-flow pump (or a bailer when purging was not possible) before sample collection.

The complete laboratory parameter set covered major cations and anions (Ca, Mg, Na, K, NH4, SO4, Cl, HCO3, NO3), Si, Al, the seven PTEs above, and total hydrocarbon content (light and heavy fractions, expressed as n-hexane), following approved standard analytical methods. Charge-balance error was calculated for all samples and accepted if below 10%; one sample (APF140) exceeds the threshold but is rescued when its Fe content (1583 µg/L) is accounted for as Fe(2+). Hydrochemical facies were defined using Piper and Stiff diagrams; PTE distributions were tested for normality and lognormality using D’Agostino tests (N >= 50) or Shapiro-Wilk tests (N < 50). Trend analysis used the Mann-Kendall test with Sen’s slope estimator on randomly selected yearly samplings.

Pre-selection of MSs followed the Italian guideline criteria: nitrate < 37.5 mg/L (75% of the 50 mg/L drinking-water TV), ammonia < 0.375 mg/L (75% of the 0.5 mg/L TV), and total hydrocarbon < 350 µg/L. Only APE040 (BPA) exceeded the nitrate threshold and was eliminated. NBL percentiles were calculated both on the per-MS representative dataset (medians where N > 1) and on the full pooled measurement dataset, with the latter selected for the final NBL proposal because the former discards half of the available data and obscures the high concentrations that can naturally arise.

Implications

Certification: Do not use this source in HMTc product or ingredient occurrence pools. The matrix is groundwater used for natural-background level characterisation under Italian/EU regulatory frameworks, not drinking-water supply, food, or any consumer-product matrix. The Cr, Cr-VI, Ni and Mn values exceed Italian drinking-water TVs but are attributed to natural water-rock interaction in serpentinite, not to anthropogenic contamination, and the proposed NBLs were accepted by the Italian authorities as a preliminary assessment.

App: Useful as environmental-pathway context for naturally elevated Cr, Cr-VI, Ni and Mn in groundwater from ultramafic and serpentinitic geological settings, including mine-impacted aquifers where neutral mine drainage rather than acid mine drainage governs PTE mobility.

Courses: Useful for teaching how to derive NBLs in geologically peculiar settings, how serpentinite mineralogy controls Cr/Cr-VI/Ni mobility, how Mg-HCO3 to Mg-SO4 facies evolution accompanies neutral mine drainage, and why median-based representative values can be too conservative when the target is the highest naturally achievable concentration rather than the typical exposure.

Wiki pages this source may touch

• chromium
• chromium-hexavalent
• nickel
• manganese
• iron
• cobalt
• zinc

Verification notes

This page was built from the full PDF, including the abstract, study area description, hydrogeology section, materials and methods, all results subsections (hydrochemical facies, pre-selection, trend analysis, spatial analysis per element, statistical distribution, dataset assessment), the NBL evaluation, comparison with similar hydrogeochemical settings, Tables 1-9, and the conclusions. Products and ingredients are intentionally empty because no food, crop, ingredient, drinking-water-supply, or consumer-product sample was analysed. Matrices use non-food descriptive slugs (groundwater, mine-impacted-groundwater, serpentinite-aquifer) consistent with the posture established for prior non-food environmental sources (e.g., son2021-nakdong-weir-sediment-metals, deng2021-mg-corncob-biochar-cu-pb-cd-remediation); the routing-audit advisory row in routing_malformed.csv is the correct surfacing channel for non-food scope. The instrument vendor and model for ICP-OES and UV-VIS are not stated in the published paper, so they are not added to the Methods section; the HACH HQ40D probe and HACH Digital Titrator are named per the Part 12 brand-firewall Exception 2 (scientific-method vendor/material names, locked 2026-05-17).

Missing-regulation-slug proposals (surfaced for Karen, not auto-created per Part 10 regulation Step 0 Lock): this paper is governed by (a) the Italian groundwater quality thresholds (TVs) in D.Lgs. 152/2006 Allegato 5 alla Parte Terza (Cr 50, Cr-VI 5, Co 50, Ni 20, Mn 50, Fe 200, Zn 3000 µg/L) and the Italian NBL guidelines [22, 33] referenced throughout, and (b) the EU Groundwater Directive 2006/118/EC and its parent EU Water Framework Directive 2000/60/EC. Neither has a wiki regulation page yet. The page intentionally does not wikilink to either to avoid creating orphans; the routes can be added when Karen approves regulation-page creation.

2026-06-03 audit application (subagent verdict REVISE, 2 findings applied, 0 rejected):

• Check 1 ⚠️ Table 4 column-shift caveat overstatement. Audit subagent flagged that the Verification-notes paragraph claiming the printed Table 4 has “Cr-VI mean (6.979) and Cr-VI median (3.1) shifted relative to the Co column header” overstates the case and reads the Co row in a way that does not match the source. Verified independently by re-reading PDF p.13 Table 4 and its surrounding text: the source’s own narrative on p.13 already says “Concerning Co, the statistics are not relevant as this element shows only two measurements > LOD, both from the same MS”, which is the correct anchor for the Co caveat. Rewrote the Table 4 caveat to cite this source statement directly and removed the column-shift speculation. The Co=not-quantifiable conclusion is preserved; the disputed mechanism explanation is removed.

• Check 2 ⚠️ Two regulation wikilinks not in taxonomy snapshot. Audit subagent flagged that [[regulations/italian-groundwater-quality-thresholds]] and [[regulations/eu-groundwater-directive-2006-118-ec]] were listed under “Wiki pages this source may touch” but neither slug exists in wiki/regulations/ and neither is in the taxonomy snapshot. Verified by listing wiki/regulations/: confirmed both are absent (existing groundwater-adjacent regulations are EPA arsenic MCL and EU drinking-water/contaminants regs only). Removed both wikilinks from the page body and added a “Missing-regulation-slug proposals” subsection in Verification notes surfacing the two regulations (D.Lgs. 152/2006 Allegato 5 and EU Groundwater Directive 2006/118/EC) for Karen’s review per Part 10 regulation Step 0 Lock.

Page history

The five most recent substantive edits to this page. The full version history lives in git; when DOI minting comes online (see schema docs), each entry below will also link to a version-pinned DataCite DOI.