Department Urban Water Management

MS2field

Real world application II

June 2019. A 70km-drive away from home, our mass spectrometer now measures the water quality in a small creek, directly in the field – our dream became true: “MS to field”! Again, the journey did not cause any problems and the unit was ready within a day. Sometimes, we fight with occasional difficulties of the bypass pump to provide a continuous sample stream reliably – in the next hardware update, we have to solve this issue.  

Real world application I

February - May 2019. With excitement, we started our first journey! Despite a few kilometres on the road, there were no incidents and the trailer was set up within half a day, ready to measure after the bake-out. Particulate matter in the raw wastewater caused recurring problems with the bypass pump, which we solved with an optimized, automated back-flushing routine. Subsequently, measurements ran smoothly and the remote process control interface that we implemented proved to be extremely helpful.  

During this deployment, our unit measured almost 2000 raw wastewater samples, originating from before and after the sand trap in the wastewater treatment plant’s influent. This implied bi-weekly cleaning, particularly of the filter unit. You can follow Michele during one of his typical maintenance days in Fehraltorf on higgs (German, with embedded Instagram posts in English) – thank you also Philipp and Benedikt for your support.

The subsequent analysis of treated effluent required almost no maintenance and we felt the automated measurements could have continued forever. We thank the open-minded, friendly staff from the wastewater treatment plant Fehraltorf for having hosted our first real-world application, and the urban water observatory team for their continued engagement.

A first evaluation of over 30 target compounds in the influent revealed almost as many distinctly different temporal patterns; thank you Pascale for your tremendous work during your semester thesis together with us. While some of these patterns observed at 20-minute resolution are plausible, others were really unexpected. Many of us probably would have struggled in accurately drawing a typical predicted pattern at this temporal resolution for most substances. We are looking forward what the non-target analyses will reveal …

Field test

December 2018. We achieved another milestone: we can now operate our mobile unit for the first time in the (almost) ‘final’ setup in the real world. Surprisingly(?), the inlet, pipe and pump to deliver the source water into the trailer requires most attention … our running gag is “It seems easier to pump ions through an Orbitrap than pumping water into the trailer”. At least no driftwood got caught at the inlet pipe in the creek (see animated sequence below).

Short sequence with driftwood dummy: 

Finishing off & roll out

November 2018. Electrical wiring of sensors, checking the PLC (programmable logic controller) connecting everything and coding rules took most of the time this month. On Friday 24 November, we celebrated the roll out (see video below). We are again very thankful for the continued support by Richi, Marco and Marc (installation solutions), Adriano (PLC), Mathias and Christian (IT and electronics) and Andy, Seba and Kay (workshop) and many others. Thank you guys for your availability and providing so promptly all the big and small things we ask for! 

 

 

Short video of the roll out:

Transfer to trailer

October 2018.  We thought we had thought of everything. For a variety of issues we had to think and rethink multiple times to find practical and robust solutions:

  • On the one hand, everything needs to be fixed and tightened to not move while driving.
  • On the other hand, how to make all parts easily accessible for maintenance, e.g. the filtration device in the source water bypass?
  • How to facilitate appropriate heating and cooling?
  • How to guarantee a safe atmosphere inside the trailer (solvents, nitrogen generator)?
  • How to prevent flooding of the trailer in case of pipe burst, or avoid the latter?
  • How to guarantee power supply and remote control of all safety relevant components?
  • How to minimize impact of vibrations?

We are enormously thankful for the support we received from our colleagues internally (many technical solutions and advice on design of remote control) and externally (plumbing and electricity)!

Now the MS is operating in the trailer and we are performing first prolonged measurements on water from the nearby creek (Chriesbach).

 

 

Short making-of video:

First success

May 2018.  In March and April, we carried out more preliminary tests and optimized the sample preparation. And most recently, the system succeeded in measuring more than 500 samples over a one-week period – one every 20 minutes – with minimal human interaction, resulting in a dataset of 500 raw files (35 GB, 2 x 10e6 scans). Pre-processing (peak picking and alignment) resulted in 90,000 time series that we evaluated with spectral analysis (periodograms). We identified approx. 50 interesting clusters. One example is a set of 38 profiles that show a distinct diurnal pattern MON-FRI but do not occur on SAT and SUN.

Assembling parts

February 2018.  The MS arrived in January. We assembled all required components in our experimental hall and tested the new workflow with “primary clarifier effluent”. Six months after the project start and just one month after the instrument arrived, we have a basic setup that seems to fulfil our requirements: a powerful pump to deliver a sample stream, a robust filtration unit, a versatile sample preparation unit, different options for front ends and a laboratory mass spectrometer that seems to cope well with conditions that are more variable than in a normal laboratory. Now we are performing first measurements with real sewage (“sand trap effluent”) and are looking for a suitable trailer.

First three months

November 2017.  Project start in September 2017! Many small details that are not an issue or easy to solve in a laboratory, require special attention and special solutions when aiming for operating an unattended system in a remote location over an extended period.

Our vision

2016/17.   Wouldn’t it be interesting to measure a broad suite of micropollutants at high temporal resolution to better understand their occurrence and fate? Do you know a system that can take hundreds of samples a week for multiple weeks, and do you have the capacity to analyse them in a lab? We don’t.

Therefore, we seek collaboration with industry partners and apply for discretionary funding. Together with Thermo and CTC, our ambitious goal* is to measure micropollutant concentrations at intervals of minutes to hours, and quantify them directly in the field – with the same, high quality as traditionally in the lab.

* A goal is a dream with a deadline - Napoleon Hill (1883-1970), author in the area of the new thought movement, one of the great writers on success.

Project Team

Dr. Michael StravsPostdoctoral scientistTel. +41 58 765 6742Send Mail
Heinz SingerSenior scientist / group leaderTel. +41 58 765 5577Send Mail
Dr. Christian StammTel. +41 58 765 5565Send Mail
Dr. Christoph OrtGroup LeaderTel. +41 58 765 5277Send Mail

Duration

September 2017  -  August 2019

Financing

Eawag  -  Discretionary Funding