901
© CSIRO 2014 Water Science & Technology
|
70.5
|
2014
Headspace gas chromatography with flame ionization detection (HS-GC-FID) for the determination of dissolved methane in wastewater D. J. Beale, G. Tjandraatmadja, M. Toifl and N. Goodman
ABSTRACT There is currently a need for a simple, accurate and reproducible method that quantifies the amount of dissolved methane in wastewater in order to realize the potential methane that can be recovered and account for any emissions. This paper presents such a method, using gas chromatography with flame ionization detection fitted with a GS-Gas PRO column coupled with a headspace auto sampler. A practical limit of detection for methane of 0.9 mg L1, with a retention time of 1.24 min, was obtained. It was found that the reproducibility and accuracy of the method increased significantly
D. J. Beale (corresponding author) G. Tjandraatmadja M. Toifl N. Goodman Land and Water, Commonwealth Scientific and Industrial Research Organisation (CSIRO), PO Box 56, Highett VIC 3190, Australia E-mail:
[email protected]
when samples were collected using an in-house constructed bailer sampling device and with the addition of 100 μL hydrochloric acid (HCl) and 25% sodium chloride (NaCl) and sonication for 30 min prior to analysis. Analysis of wastewater samples and wastewater sludge collected from a treatment facility were observed to range from 12.51 to 15.79 mg L1 (relative standard deviation (RSD) 8.1%) and 17.56 to 18.67 mg L1 (RSD 3.4%) respectively. The performance of this method was validated by repeatedly measuring a mid-level standard (n ¼ 8; 10 mg L1), with an observed RSD of 4.6%. Key words
| dissolved methane, headspace gas chromatography, method development, sludge, wastewater
INTRODUCTION Methane is a non-toxic and non-poisonous gas. For wastewater operators and managers, dissolved methane is considered to be an opportunity and/or a liability (Tauseef et al. ). The opportunity lies in the capture of methane, where it can then be used as a biofuel in energy production. Conversely, if the methane is not captured, there is a potential liability when methane is released into the atmosphere where its impact as a greenhouse gas is both well studied and documented (Cloy & Smith ). Furthermore, in today’s growing and changing ‘carbon’ economy, the pressure for water and wastewater utilities to account for and offset fugitive emissions is increasing (Beale et al. ). As such it is important for wastewater utility operators to quantify methane balances in wastewater treatment facilities in order to get a better understanding of the potential energy that can be produced and the amount of methane lost to the atmosphere. The measurement of dissolved methane requires the combination of representative sampling and accurate doi: 10.2166/wst.2014.298
analytical methodologies. Given the low solubility of methane at ambient temperatures, sampling methods need to minimize disturbances in order to reduce losses of dissolved methane to the atmosphere. This is a broad challenge facing the sector, as the confidence in the accuracy and reliability of current approaches is lacking. In order to address this challenge, the aim of this research is to develop a robust methodology for the collection of wastewater and develop a protocol that ensures a high, accurate and representative recovery of methane for analysis by headspace gas chromatography with flame ionization detection (HS-GC-FID). As such, this paper is structured in the following way. First, a review of the various sampling approaches and the analytical techniques for measuring dissolved methane in aqueous solutions is given. Second, a description of the sampling method and analytical conditions applied is then provided. Lastly, the key findings of this research are presented, and preferred analytical methodologies are discussed.
902
D. J. Beale et al.
|
HS-GC-FID methane recovery
Current sampling and analysis of dissolved methane A large number of sampling devices and analytical methods have been adopted for sampling dissolved gases in groundwater, surface waters and wastewater (Hirsche & Mayer ; Loughrin et al. ). These techniques typically involve grab samples collected manually, with the aid of pumps and/or the use of flow cells and membranes for selective gas diffusion (Kampbell & Vandegrift ; Hirsche & Mayer ; Loughrin et al. ; Lomond & Tong ; Liotta & Martelli ). In situ sampling in the field and in confined spaces, such as a wastewater treatment facility, is generally conducted by one of two means: collection with devices that are lowered into a well, such as bailers, and retrieval of water to the surface using containers or displacement pumps operating at low flow rates (Walsh & McLaughlan ; Souza et al. ). Only a few studies have conducted comparisons of these sampling techniques: Walsh & McLaughlan () compared the use of open and closed bailers with two different pumps for sampling of water from wells. No significant differences in accuracy were observed for the dissolved methane concentrations detected using the different sampling techniques; however, the precision (recorded as relative standard deviation (RSD)) was greater for the bailers operated by skilled users (Walsh & McLaughlan ). Once collected, the current preferred method for many utilities for the measurement of dissolved methane in wastewater is based on the US EPA method, titled ‘Technical guidance for the natural attenuation indicators: methane, ethane and ethene’ (US EPA ). The US EPA method is based on work by Kampbell & Vandegrift () and the analysis of dissolved hydrocarbons in groundwater. The US EPA method determines the dissolved gas concentration using Henry’s Law, where it is assumed that the concentration of a gas in the liquid phase is proportional to the partial pressure of the gas above the liquid. However, wastewater is a very different sample matrix to that commonly observed for groundwater samples, for which the US EPA method was developed. As such, the application of this method in its current form has raised some questions about the validity of this approach. With this in mind, researchers have been investigating variations to the US EPA method on a range of sample matrices, namely: wastewater, seawater, brackish water and waters with a high ionic strength (Alberto et al. ; Gal’chenko et al. ; Lomond & Tong ; Souza et al. ; Jahangir et al. ; Liotta & Martelli ; Daelman et al. ). Review of the literature indicates that the salting out method appears to provide more reliable
Water Science & Technology
|
70.5
|
2014
results with a much simplified methodology that eliminates the use of Henry’s Law calculations. In addition, as the focus of the analysis is on wastewater samples that comprise relatively high concentrations of methane when compared to groundwater samples, the use of an inert gas to prepare the headspace seems unnecessary. The amount of methane in the air within the headspace would be negligible in comparison to the methane dissolved in the wastewater sample. In addition to developing methodologies for different sample applications, other method parameters have also been investigated and modified: for example, variations of headspace volume, temperature, agitation time/sample equilibrium conditions and salinity.
METHODS Wastewater sampling method Two techniques were selected for sampling wastewater from the inlet well to a pumping station located after an anaerobic pond at Melbourne Water’s Western Wastewater Treatment Plant (Australia): a sampling bailer device and a peristaltic pump with an inverted bottle. The wastewater samples were analyzed for chemical oxygen demand (COD) and total dissolved solids (TDS), which were ca. 500 mg L1 and 840 mg L1 respectively. Sampling method 1: bailer Briefly, the bailer lowered a 100 mL Schott sampling bottle to a pre-selected depth of 1 m, at which depth the bailer was uncapped. After filling, the bottle was capped and returned to the surface. The bailer was constructed from a polyvinylchloride tube of 65 mm diameter (85 mm maximum at joints) and was adjustable in length (maximum 4 m). It had an internal rod that allowed the opening and closing of the sample bottle. To aid the analytical testing, the sample bottle cap had a flexible silicon/polytetrafluoroethylene (PTFE) septum allowing the use of syringes for sample removal and analysis. Sampling method 2: peristaltic pump A peristaltic pump (Masterflex L/S Digital Drive, 600 rpm, 115/230 VAC) was used to collect wastewater at a low suction rate (67 mL min1, then increased to 133 mL min1 to avoid bubble formation in the suction line). The suction inlet was lowered to a pre-selected depth of 1 m using a guiding pole
903
D. J. Beale et al.
|
HS-GC-FID methane recovery
Water Science & Technology
and the wastewater was pumped into sample vials. Two techniques were adopted: wastewater was pumped into an inverted 100 mL Schott bottle and filled with wastewater until 2× the volume of the bottle was transferred, and secondly, wastewater was pumped directly into 22 mL vials. The outlet of the tube was placed against the inside wall of the glass vial in order to avoid agitation. After filling, the sample bottle and vials were capped/crimped. Wastewater sludge For the purpose of method development, wastewater sludge samples were collected from a sampling port on the side of an anaerobic digester at the Melbourne Water Eastern Treatment Plant (November 2013). Wastewater sludge samples were selected for method development as they represented samples that would comprise the largest proportion of solids and dissolved methane. The COD and TDS of the wastewater sludge samples were ca. 5,450 mg L1 and ca. 3,500 mg L1 respectively.
|
70.5
|
2014
sampling bags were purged with helium prior to blending the desired methane concentrations; all standards were produced to 1 atmosphere. A seven-point calibration curve was created establishing method linearity and reporting limits. Twenty-one headspace vials were prepared with MilliQ water (with the addition of 25% NaCl and 100 μL HCl and sonicated for 30 min). Known quantities of methane were then injected using a gas-tight syringe through the septum into the water of the vials containing MilliQ water, attaining concentrations of methane in the range of 0.5 to 25 mg L1. Headspace gas chromatography with flame ionization detection Samples were analyzed using a headspace auto sampler (Agilent Headspace Auto Sampler 7694E) coupled with an Agilent 6890N gas chromatogram (GC) equipped with a 30 m × 0.320 (mm) wide bore GS-Gas Pro column (Agilent Technologies, Mulgrave) with FID. The headspace heating zone was maintained at 55 C, with a headspace loop and transfer line of 105 C and 135 C respectively. The samples were agitated for 10 min prior to injection (500 μL). The carrier gas was high purity helium with a flow rate of 5 mL min1. The oven was programmed with an initial temperature of 40 C for 1 min, increasing at 15 C min1 to 150 C, and then held for 1 min. The injector was set at 200 C. The FID detector was set at 240 C. The FID hydrogen flow rate was 40 mL min1, the air flow rate was 450 mL min1, and nitrogen makeup gas flow rate was 32 mL min1. GC ChemStation (Agilent Technologies, Rev A.10.02) was used for signal acquisition and peak integration. W
W
Sample transport After collection, capped samples were labeled, inverted upside down to avoid the loss of any gases, inserted in plastic bags and placed in a cooler filled with ice for transport. Upon arrival at the laboratory, samples were stored at
