Source: WASTE HUB, LLC submitted to NRP
VALUE EXTRACTION AND PROTEIN RECYCLE FROM LIQUID ACID WHEY USING CATALYTIC ELECTROLYSIS
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
1009787
Grant No.
2016-33610-25355
Cumulative Award Amt.
$100,000.00
Proposal No.
2016-01081
Multistate No.
(N/A)
Project Start Date
Jul 1, 2016
Project End Date
Apr 30, 2018
Grant Year
2016
Program Code
[8.5]- Food Science & Nutrition
Recipient Organization
WASTE HUB, LLC
5796 S KITTREDGE CT
AURORA,CO 80015
Performing Department
(N/A)
Non Technical Summary
Greek (strained) yogurt represents nearly half of the $5B U.S. yogurt industry and has grown at a rate of more than 60 percent per year for the past eight years. For every pound of milk used in the manufacture of Greek yogurt, only one-quarter to one-third ends up in the final product. The remaining waste is liquid acid whey, which contains more than 93 percent water, along with minerals, protein, lactose-based compounds. Acid whey is environmentally hazardous and there are few cash-positive options for disposal or reuse. Waste Hub LLC, under an exclusive sublicense from Catalytic Innovations LLC, has filed a patent application for an electrolysis process that will recover neutral water and separated proteins for recycle while generating fuel-cell electricity and pure carbon dioxide for sale. The process will use the award-winning, heterogenized, iridium-based, molecular oxidation catalyst, developed at Yale University, to selectively oxidize the organics using a fraction of the energy required for state-of-the-art electrolysis processes. Waste Hub will assemble and test a solids-separation and electrolysis unit to refine the Waste Hub process and verify technical and economic viability. We will further develop physics-based predictive models and use them for full-scale system design, economic forecasting, and preparation for a Phase II pilot with a major producer of Greek yogurt. This application is in response to Topic Area 8.4 - Air, Water, and Soils and addresses the cross-cutting USDA priority of Energy Efficiency and Alternative and Renewable Energy from agricultural waste.The hypothesis behind this project is that electrolysis using an iridium-based, heterogenized molecular oxidation catalyst on the anode and a platinum catalyst on the cathode will readily and durably oxidize the organic content of liquid acid whey with relatively low overpotential and will comprise an economically-sound alternative energy technology for processing liquid acid whey. The technical objectives are to confirm this hypothesis, namely:To demonstrate experimentally that acceptable rates of reaction and current density can be achieved using a continuous feed of liquid acid whey with low potential (<2 V).To determine the most cost-effective method of separating suspended solids (protein) and asses the ability to recycle the protein to the processTo collect samples and analyze the flow rates and composition of all outlet streams (H2, O2/CO2, water, solids) and use this information in the development and validation of a process simulation model of the entire Waste Hub system, including the fuel cell and CO2 liquefaction process.To produce designs and economic assessments of a scaled-up pilot and a full-scale system using the validated process model.The key research questions to be addressed in this Phase I project are as follows:What method of solids separation maximizes separation and prevents electrolysis performance degradation while preserving and recovering solids for reuse?What is the effect of solids loading on performance (outlet streams, durability, energy consumption) of the electrolysis system?What is the effect of applied potential on performance (outlet streams, durability, energy consumption) of the electrolysis system?Based on Phase I results, what would be the design, mass balance, energy balance, cost, and financial return of a pilot-scale Waste Hub system to be implemented in Phase II at Stonyfield, and of a complete commercial-scale Waste Hub system at 25,000 gal/day of liquid acid whey?Waste Hub will use samples of liquid acid whey provided by Stonyfield Farm in Londonderry, NH to conduct experiments in 1) suspended solids and protein separation and 2) electrolysis of filtered and non-filtered acid whey in which the anode catalyst, membrane material, and applied voltage will be varied to determine the operating conditions for best performance. Composition of inputs and outputs will be measured in each experiment, including feed and filtered acid whey in separation experiments and feed acid whey, anode gas output (CO2/O2), cathode gas output (H2), and effluent water. The experimental conditions and composition data will be used as inputs for a system design model to be developed in parallel.The system design model will be a relatively simple, steady-state, semi-empirical model with parameters adjusted within reason to achieve agreement with experimental data. We will take into account mass transfer from the fluid to the electrodes in the electrode chambers and the equilibrium chemistry of the half-reactions taking place at the electrodes. The model should yield reasonable estimates of power consumption, gas compositions leaving the electrodes, and organic composition of effluent water. We will not attempt to address the minor, non-reacting constituents of acid whey other than to track their concentrations through the process based on changes in water and organics content. MS Excel will be used to develop the model in Phase I and will be upgraded or integrated to a more sophisticated process modeling package (i.e. ChemCAD, ASPEN) in Phase II.
Animal Health Component
40%
Research Effort Categories
Basic
20%
Applied
40%
Developmental
40%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
1110210202090%
1330199202010%
Goals / Objectives
Waste Hub LLC is a Colorado small business whose mission is to take a proactive role in creating valuable alternatives to disposal for industrial waste streams. In cooperation with Stonyfield Farm, Yale University, and Catalytic Innovations LLC, Waste Hub is developing a breakthrough electrolysis technology to address a national challenge in the dairy industry. We have demonstrated at laboratory scale that catalytic electrolysis may be used effectively to convert what is now a high-volume, expensive, and environmentally hazardous liquid acid whey stream into recyclable water and protein, bio-derived hydrogen energy, and salable liquefied carbon dioxide. Greek yogurt is a rapidly-growing segment of the dairy industry, providing a healthy food option that is consumed in high volume by the public. However, the long-term sustainability of Greek yogurt manufacture is challenged by the generation of two to three pounds of liquid acid whey waste for every pound of yogurt product. Disposal is expensive because of the acidity (pH between 4 and 5) and high biological oxygen demand (BOD) of acid whey. Further, the options for productive use are limited and few if any of the available options produce a positive economic return for the manufacturer.Waste Hub's application is in response to Research Topic 8.4 Air, Water, and Soils and NIFA National Challenge Area 6 (Water) because we are developing a technology that has the commercial objective of cost-effectively treating and recycling enormous amounts of hazardous wastewater in the dairy industry. This application also addresses the USDA cross-cutting priority of developing alternatives to fossil fuels and the reduction of energy consumption by converting the hydrogen content of the organic fraction in acid whey to electricity via a fuel cell.The Waste Hub team has developed patent-pending technology that uses electrolysis to treat liquid acid whey (Sheehan 2015), neutralizing and removing organic contaminants from the waste stream. This is coupled with an exclusive sublicense from Catalytic Innovations to a patent-pending electrolysis catalyst chemistry that was developed at and is owned by Yale University (Sheehan et al. 2014). This technology, published in March of 2015 in Nature Communications (Sheehan et al. 2015), currently holds the world record for efficiency in electrolytic water and hydrocarbon oxidation processes. It has also resulted in finalist nominations for the Institute of Chemical Engineers Global Energy Award and the Collegiate Inventors Competition. In this Phase I SBIR project, Waste Hub will develop a physics-based design model and assemble, test, and demonstrate a pilot-scale electrolyzer using liquid acid whey waste from the Stonyfield Farm Greek yogurt manufacturing plant in Londonderry, New Hampshire. While we will only sample and discard the outlet streams from the process in this Phase for purposes of design and development of a complete pilot in Phase II, the ultimate objectives of the technology are to provide a cash-positive technology option for Greek yogurt manufacturers that will 1) generate electricity from the hydrogen using a PEM fuel cell, 2) separate oxygen and liquefy the CO2 stream for sale, and 3) recycle pre-filtered solids and clean effluent water to the yogurt manufacturing process.The hypothesis behind this project is that electrolysis using an iridium-based, heterogenized molecular oxidation catalyst on the anode and a platinum catalyst on the cathode will readily and durably oxidize the organic content of liquid acid whey with relatively low overpotential and will comprise an economically-sound alternative energy technology for processing liquid acid whey. The technical objectives are to confirm this hypothesis, namely:To demonstrate experimentally that acceptable rates of reaction and current density can be achieved using a continuous feed of liquid acid whey with low potential (<2 V).To determine the most cost-effective method of separating suspended solids (protein) and asses the ability to recycle the protein to the processTo collect samples and analyze the flow rates and composition of all outlet streams (H2, O2/CO2, water, solids) and use this information in the development and validation of a process simulation model of the entire Waste Hub system, including the fuel cell and CO2 liquefaction process.To produce designs and economic assessments of a scaled-up pilot and a full-scale system using the validated process model.The key research questions to be addressed in this Phase I project are as follows:What method of solids separation maximizes separation and prevents electrolysis performance degradation while preserving and recovering solids for reuse?What is the effect of solids loading on performance (outlet streams, durability, energy consumption) of the electrolysis system?What is the effect of applied potential on performance (outlet streams, durability, energy consumption) of the electrolysis system?Based on Phase I results, what would be the design, mass balance, energy balance, cost, and financial return of a pilot-scale Waste Hub system to be implemented in Phase II at Stonyfield, and of a complete commercial-scale Waste Hub system at 25,000 gal/day of liquid acid whey?
Project Methods
Investigation of Solids SeparationThe objective of this task is to evaluate cross-flow filtration, settling, and flocculation as potential methods for solids separation.Cross-Flow Filtration: We will use a low-pressure pump to deliver 1 gal/hr of raw liquid acid whey from a stirred storage beaker through a cross-flow membrane filter holder with three different grades of membrane/filter to filter solids for 30 minutes each. The pressure upstream of the filter will be monitored to assess plugging of the filter over time. Samples of fluid downstream of the filter will be taken initially and then every ten minutes. After each test, the filter will be removed and weighed to determine the amount of solids separated. The solids will then be removed, washed, and pressed with 250 ml of deionized water. The following data will be collected and analyses conducted for each test:pH, total suspended and dissolved solids (ASTM D5907-13), total organic carbon (ASTM D7573-09), total protein (Kjeldahl or Dumas method), and true protein (Kjeldahl or Dumas method with protein precipitation using trichloroacetic acid) of input and output acid whey samples.Upstream pressure vs timePercent solids removed by mass over the 30 minute testSettling: Two stainless steel settling containers will be arranged in series such that the first receives liquid acid whey at a rate of 1 gal/hr and transfers liquid via overflow to a second, equally-sized beaker. Fluid will be withdrawn from the top of the second beaker at 1 gal/hr for 30 minutes. Three tests will be conducted with three different beaker sizes to examine the effect of residence time. Samples downstream of the second beaker will be taken initially and then every ten minutes. The settled solids in each of the two beakers will be isolated at the end of each test by gently withdrawing the fluid from each beaker with minimal disruption of the solid layer. The solids will then be washed and pressed with 250 ml of deionized water. The following data will be collected and analyses conducted for each test:pH, total suspended and dissolved solids (ASTM D5907-13), total organic carbon (ASTM D7573-09), total protein (Kjeldahl or Dumas method), and true protein (Kjeldahl or Dumas method with protein precipitation using trichloroacetic acid) of input and output acid whey samples.Percent solids removed by mass over the 30 minute testFlocculation: We will use a low-pressure pump to deliver 1 gal/hr of raw acid whey from a stirred beaker into a stainless steel tank along with a pH-metered flow of dilute sulfuric acid. Precipitated proteins will be removed from the tank by skimming or filtration. The solids will then be washed and pressed with 250 ml of deionized water. The following data will be collected and analyses conducted for each test:pH, total suspended and dissolved solids (ASTM D5907-13), total organic carbon (ASTM D7573-09), total protein (Kjeldahl or Dumas method), and true protein (Kjeldahl or Dumas method with protein precipitation using trichloroacetic acid) of input and output acid whey samples.Percent solids removed by mass over the 30 minute testExperimental Research and Testing of ElectrolysisPreliminary proof-of-concept experiments show that, both with and without removal of solids from the liquid acid whey sample, an anode comprised of the molecular iridium oxidation catalyst supported on tin oxide-based mesoporous supports are capable of neutralizing liquid acid whey by oxidizing organics to CO2 and reducing water to H2 as shown by the catalytic currents in Figure 2. In these small lab-scale experiments, any solids present in the waste stream float to the top of the anode chamber by electroflotation. In similar schemes using small PEM electrolyzer test cells, the membrane was found to be functional in the presence of a filtered liquid acid whey sample. With these successful proof-of-concept experiments in mind, the objective of this task is to scale-up and optimize the components present in a small pilot-scale electrolyzer. As a part of this optimization process, the stability of each component will be tested rigorously, optimum operational parameters will be explored, and purity of the product streams will be determined.The electrochemical device contains an anode and cathode, each in its own respective chamber and separated by a bipolar membrane. Upon application of an electric potential from a DC rectifier, CO2 gas is generated from the oxidation of organic matter at the surface of the anode, while H2 gas is generated by the reduction of water at the cathode. Due to the overpotential required to achieve high rates of CO2 evolution, there will be some residual water oxidation at the anode, and as a result some O2 contaminating the CO2 stream. Determining the ratios of CO2 to O2 and the effect of the applied electric potential on those ratios is one of the goals of this project. Additionally, as demonstrated in the proof-of-concept experiments, the gas bubbles generated at the anode serve to float any remaining whey solids to the surface of the anode chamber, where they can then be skimmed and collected. Between the anode and cathode chambers, a bipolar membrane serves to spatially separate the output streams, and provides a channel by which the effluent water can be expelled from the electrolyzer. Since the majority of the liquid acid whey stream is water, the highest percent by weight output will also be this treated water. In order to scale-up and optimize this process for the pilot scale, the following tests must be performed:Nine 1-hour tests monitored using chronoamperometry [3 applied potentials, 3 solids loading (0, 0.75%, 1.5%) to determine to what extent electroflotation alone will it remove solids]Flow rate: 1 gal/hrMonitor flow rates of H2, O2/CO2, water over timeMonitor power consumption over timeSample collection every 15 minutes (H2 stream, O2/CO2 stream, water stream)H2 stream sample analyses (gas chromatography)O2/CO2 stream sample analyses (gas chromatography)Water stream sample analyses (pH, TSS, TDS, TOC, Total Protein, True Protein)Screen commercially available bipolar membrane technologiesExtended tests (>8 hours) under the best conditions identifiedProcess Modeling, Design and EconomicsIn this Phase I project, we will develop a mathematical model of the electrolysis process and validate predictions using the data concerning inputs and outputs generated in the electrolysis experiments. We will also use the model to interpret the results and optimize the process through parameter studies. In this Phase, we will develop a relatively simple, steady-state, semi-empirical model with parameters adjusted within reason to achieve agreement with experimental data. We will take into account mass transfer from the fluid to the electrodes in the electrode chambers and the equilibrium chemistry of the half-reactions taking place at the electrodes. The model should yield reasonable estimates of power consumption, gas compositions leaving the electrodes, and organic composition of effluent water. We will not attempt to address the minor, non-reacting constituents of acid whey other than to track their concentrations through the process based on changes in water and organics content. MS Excel will be used to develop the model in Phase I and will be upgraded to a more sophisticated process modeling package (i.e. ChemCAD, ASPEN) in Phase II.

Progress 07/01/16 to 02/28/17

Outputs
Target Audience:The primary target audience for this project is the dairy industry, specifically Greek yogurt production plants and facilities and their respective ecosystems. The ecosystem around acid whey remediation and reuse includes waste management contractors, R&D teams at the plant searching for new applications of acid whey, and prospective buyers of the value-added products made from acid whey. The secondary audience for this project is the electrolysis and fuel cell industries. This market segment has been plagued with low profitability over the course of the last several decades. The infrastructure for a hydrogen economy has not been developed due to the low cost of fossil fuels, and profitable, value-added applications of energy efficient polymer electrolyte membrane electrolysis technologies have been limited. Use of this technology for wastewater remediation is a new angle that brings potential value to this industry without relying on the price of hydrogen derived from water splitting versus hydrogen derived from much cheaper natural gas sources. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?For the two scientists working full-time on the project, Stafford W. Sheehan and Chi Chen, this has been a training experience in practical R&D, having both completed their PhD degrees in chemistry within the last two years. This includes evaluating cost structure, rapid screening of materials and components, constructing prototypes, and systems evaluation at an engineering level How have the results been disseminated to communities of interest?These results have been disseminated to Stonyfield Farms and Groupe Danone in e-mail communications and meetings throughout the project, satisfying our primary target customer. To the electrolysis community at-large, preliminary results have been published in Industrial & Engineering Chemistry Research, an American Chemical Society research journal. These have also been disseminated and discussed in meetings with PEM electrolysis industry leaders Proton Onsite and Axine Water Technologies. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? The key research questions outlined above were answered by performing approximately 100 electrolysis experiments on the prototypes we constructed, which included 14 day-long stability tests complete with successful collection and NMR and GC-MS evaluation of the output gases and liquids during the performance period of this grant. This was accompanied by nearly 50 protein assays on acid whey at various points after its production (fresh from the production line out of the production facility, after freezing and thawing, pre- and post-filtering) to determine how variation in the input wastewater stream effects the electrolysis process, and evaluate value-added methods of filtration that can be added prior to electrolysis to improve the economics of this process. We also investigated the basic science behind the results that we gathered in our prototype using standard 3-electrode electrochemical techniques, to understand why and how our catalytic electrolysis system operated in the manner it did. With respect to the specific research questions: 1. We determined that 5kD diafiltration maximizes electrolysis performance while minimizing membrane degradation, while leaving a protein-rich retentate. Over the course of evaluating filtration methods and membranes with the subcontractor Smartflow, we evaluated both nanofiltered membranes and diafiltered membranes, as well as performed ELIZA tests on the retentate to determine relative and absolute concentration of both alpha-lactalbumin and beta-lactoglobulin. We found that the acid whey retentates possess a higher concentration of alpha-lactalbumin than expected or found previously in literature. This is a critical ingredient in baby formula and opens a pathway to additional economic incentive for our system, by allowing the dairy farms that implement our solution to adopt a new product based on the high alpha-lactalbumin content. 2. We found that large solids can be tolerated by an electroflotation prototype, however a PEM prototype does not operate stably while solids are present. In investigating both systems, however, we found that the electroflotation prototype employing dimensionally stable electrodes does not possess sufficient ionic conductivity to function in the acid whey wastewater. This low ionic conductivity between electrodes necessitates that the electroflotation prototype operate at high potentials, which causes rapid deactivation of the catalysts. Our hypothesis is that this may be due to oxidation of the underlying titanium anode support under high potential conditions, building an insulating TiO2 layer underneath the iridium-based electrocatalysts on the anode. Because the solid-tolerant electroflotation prototype encountered this hurdle, we moved forward with the PEM prototype - therefore, solids must be removed. 3. We found that applied potentials below 2.0 V and 2.5 V are not sufficient to achieve the rates required for wastewater remediation, and that the polymer electrolyte membranes do not survive without fouling at these low potentials. Potentials higher than 4.0 V oxidized amino acids into ammonium hydroxide, which served to deactivate the sulfonic acid groups in the Nafion membrane. Thus, we have determined the "sweet spot" for PEM wastewater treatment is between 3.0 V and 3.5 V, and we ran our further experiments at 3.5 V. While the iridium-based catalysts that we performed our benchtop proof-of-concept experiments on were functional and stable at potentials below this, we found that the iridium-based catalysts did not possess the same activity when mounted on a membrane in a PEM electrolyzer. Our data both on the acid whey being fed into the system and on the treated water coming out suggests that this is because the iridium-based catalysts oxidize amino acids present in the acid whey waste stream to form ammonium hydroxide, which poisons the sulfonic acid groups in the polymer electrolyte membrane separating the anode and cathode of the electrolyzer. This, we hypothesized, causes the membrane to deactivate. The same effect was seen generally when there was a high nitrogen content in the acid whey samples received from Stonyfield Farms in Kjeldahl nitrogen tests, which provided further evidence for our hypothesis. While the iridium-based catalysts we performed our proof-of-concept experiments outside our prototype were not functional, we did find that platinum-based catalysts did not suffer from the same issue. These catalysts operate by oxidizing organic contaminants with hydroxyl radicals, which may possess lower selectivity for the amino acids that are oxidized by iridium-based catalysts. 4. With the results we gathered in Phase 1, we are able to construct a preliminary mass balance and energy balance for the electrolysis system that we plan to build in Phase 2. With this data, we have a clear and straightforward path for developing our Phase 2 research plan, in order to assess the viability of this solution on the pilot scale. From the work we plan to do in Phase 2, we will be able to assess the economics of the overall solution at-scale and prove a case for construction of commercial wastewater treatment facilities. Over the course of answering these questions, we have explored the variables present in the manufacture and design of the components that would be required for large-scale deployment of this technology. Central to this is the manufacture of membrane electrode assemblies (MEAs), in which we found three scaling limitations we plan to address in the future: 1. Uniformity and control of deposition of catalyst inks onto membranes must be precise. This requires electrospray, decal printing, or other large-scale manufacturing techniques. Non-uniform catalyst layers severely inhibit overall performance and cannot be tolerated as the system is scaled. 2. MEAs must be heat-pressed in a reproducible manner so that stability of the system over the surface of the MEA is consistent and membrane deactivation is not exacerbated. 3. Commercially sourced MEAs have large batch-to-batch variability. While our experimental conditions did not change, we found that the effectiveness (both stability and activity) of MEAs bought form different industrial suppliers, and bought from the same supplier in different batches, varied. The membranes that have been studied the most and optimized in the industrial electrolysis space are Nafion 117, which necessitated its use in Phase 1 after screening the other membrane types outlined in our proposal. This introduced an additional variable of electroosmotic drag to our system, where some of the wastewater crossed the membrane during treatment due to the membrane's thinness, which we plan to address further in Phase 2. For this and future development, we initiated a partnership with electrolysis company Proton Onsite, from whom we purchased components of our prototype system so that MEA manufacture risk is minimized. Over the course of this project, we found that acid whey ferments to produce ethanol, which came as a surprise to both our team and the engineers at Stonyfield Farms. Our experiments in the PEM electrolyzer prototype were performed after filtration, which allowed some time for the acid whey to ferment and build up an appreciable ethanol concentration (>5%). We found that the acid whey batches where the ethanol was produced allowed for longer-lived MEAs. Additional experiments showed that, when ethanol was added to acid whey with lower ethanol concentration (made by monitoring acid whey custody and keeping it frozen), the stability of the membrane increased drastically. Concentration of lactose and nitrogen-containing organics also have an effect on MEA activity. One hypothesis as to why this occurs is because the ethanol is an easier oxidation target than amino acids are, thus leading to slower ammonia production and less membrane fouling. Our evaluations how this can be used beneficially for wastewater treatment are ongoing.

Publications

  • Type: Journal Articles Status: Published Year Published: 2017 Citation: Chen, C.; Bloomfield, A. J.; Sheehan, S. W. "Selective Electrochemical Oxidation of Lactic Acid Using Iridium-Based Catalysts" Ind. Eng. Chem. Res., 2017, 56 (13), pp 35603567