Low Energy Process Control January 23 rd, 2013 1
WEF Webcast Low Energy Process Control Ammonia based aeration control Leiv RIEGER, inctrl Solutions Inc., Canada Acknowledgement Peter L. Dold, EnviroSim Richard M. Jones, EnviroSim Charles B. Bott, HRSD William J. Balzer, HRSD 4 2
Overview Context Aeration control Nitrification fundamentals Aeration control strategies Control fundamentals Case studies Conclusions 5 Context Aeration costs 6 3
Context Benefits Ammonia-based aeration control WWTP Morgental 35,000 PE 3.5 mgd WWTP Thunersee 130,000 PE 10 mgd WWTP Werdhoelzli 600,000 PE 50 mgd simulation full-scale simulation full-scale simulation Energy -30% -20% -30% -16.5% -25% TN removal +48% +40% +60% +40% +32% Annual net savings $ 53 000 $ 360 000 $ 1 200 000 Rieger et al., WER 2012 7 Context Benefits Case study HRSD s Nansemond WWTP 5-stage Bardenpho, 60,000 m 3 /d (16 mgd), 250,000 PE Simulation study 8 4
Context Influent variability WWTPs are highly dynamic systems... Olsson, 2008 9 Context High variability of incoming load Fixed reactor volumes WWTP design based on peak load Unused capacities Nitrification is the rate limiting step and therefore the primary target of BNR aeration control strategies 10 5
Nitrification fundamentals Nitrification requirements Sufficient provision of dissolved oxygen Ammonia as substrate (+ essential nutrients) Sufficiently long aerobic sludge retention time Sufficient mass of nitrifiers Autotroph 11 Nitrification fundamentals DO constraints nitrification kinetics At 2 mg DO/L: ca. 80% of max. rate 12 6
Nitrification fundamentals Ammonia as substrate Typical ammonia profile from fully aerated plant 13 Nitrification fundamentals Ammonia effluent variations Typical effluent ammonia variation from fully aerated plant SRT aerob = 8 days Average 0.5 mgn/l Reasons for peaks? 14 7
Nitrification fundamentals Nitrifier mass The mass of nitrifiers changes slowly The total mass depends on average ammonia load and SRT The influent ammonia load may vary substantially over a day Ammonia break-through often: due to limited mass of nitrifiers not a problem of insufficient oxygen (or other limiting components) 15 Aeration control strategies DO versus ammonia control BOD removal PST Denitrification Nitrification FST Control handle: Aeration DO control aims for optimal DO for aerobic processes NH 4 control optimizes nitrification process 16 8
Aeration control strategies Ammonia-based aeration control 1) Limiting aeration: Reduce energy consumption, increase denitrification, improve bio-p performance 2) Reducing effluent ammonia peaks: Reduce the extent of effluent ammonia peaks 17 Aeration control strategies 1) Limiting aeration Nitrifiers grow slowly Rate limiting step Pure DO control Aeration even after ammonia is gone NH 4 control Intermittent aeration/varying intensity to limit nitrification O 2 NH 4 Tailored nitrification/denitrification 18 9
Aeration control strategies 1) Limiting aeration: Cascaded NH 4 /DO control Aeration intensity control (or intermittent aeration) Measured variable (Actual value) Pressurized air DO Controller Manipulated variable Reference variable (setpoint) NH 4 controller DO f(nh 4 ) O 2 M NH 4 19 Aeration control strategies 1) Limiting aeration: Direct NH 4 control Measured variable (Actual value) Pressurized air Reference variable NH (setpoint) 4 Controller Manipulated variable NH 4 M High NH 4 leads to over-aeration Additional DO probe More difficult to tune 20 10
Aeration control strategies 2) Reducing ammonia effluent peaks Intensity control: Manipulate aeration intensity early to create buffer for incoming peak Volume control: Change aerated volume by switching on/off swing zones 21 Control fundamentals Feedback versus Feedforward control Feedback control Setpoint Disturbances z Target variable Reference variable / Setpoint u ε Controller y Final Control Element Process Controlled variable x Measured variable r Measuring Device Measure process answer 22 11
Control fundamentals Feedback versus Feedforward control Feedforward control Disturbances z Measure process disturbance System model r Measuring Device Reference variable / Setpoint u ε Controller y Final Control Element Process Controlled variable x Fast reaction before disturbance hits the plant Process model required Must be complemented by feedback signal More sensors required 23 Control fundamentals Feedback+Feedforward control NH 4 Q Feedforward Controller Measured variable Press. air O 2 Maximumcriteria Ref. variable DO Controller Manipulated variable M NH 4 controller DO f(nh 4 ) NH 4 24 12
Control fundamentals Variable DO setpoint control NH 4 setpoint DO setpoint Airflow setpoint Valve opening Air flow NH 4 controller DO controller Airflow controller Air flow system Aerator DO NH 4 Gustaf Olsson, 2012 25 Case study HRSD s Nansemond WWTP 26 13
Case study Nansemond Influent / temperature scenarios Input 1) Dry weather conditions at average temp. of 12 C Input 2) Dry weather conditions at average temp. of 20 C Input 3) Dry weather conditions at average temp. of 30 C Input 4) Ammonia peak at average temperature of 12 C 27 Case study Nansemond Control scenarios Base case: Existing strategy: DO control CS 1: DO probes moved CS 2a: Ammonia feedback: continuous change of DO setpoint PID with DO setpoint 0.5-2 mgdo/l CS 2b: 2a but DO setpoint 0-2 mgdo/l CS 3a: Ammonia feedback: high-low / intermittent aeration On-Off with DO setpoint 0.5/2 mgdo/l CS 3b: 3a but DO setpoint 0/2 mgdo/l CS 4: Feedforward+Feedback ammonia control 28 14
Case study Nansemond Current DO control strategy DO setpoint 2.5 mg/l DO setpoint 2.0 mg/l DO setpoint 1.0 mg/l 29 Case study Nansemond Base control strategy Aeration zone 1 Aeration zone 2 Aeration zone 3 Low DO conc. in downstream section of second aeration zone Sensor information vs. DO profile 30 15
Case study Nansemond Optimal DO probe location Base control strategy Control strategy 1 (DO probes AZ1&2 moved downstream) 31 Case study Nansemond Control strategy (FF+FB) Feedforward DO Controller AAA E/F Q NH 4 O 2 M airflow O 2 airflow M airflow M O 2 NH 4 Feed-forward NH 4 high/low Controller DO Controller Aer4-7 a-e DO Controller AAA E/F Feed-back NH 4 PID Controller Feedback Selector 32 16
Case study Nansemond Feedforward control Feedforward control only activ at 12 C 33 Case study Nansemond Feedforward control Feedforward control has no significant impact on effluent ammonia 34 17
Case study Nansemond Feedforward control Even at extreme peak events limited impact of feedforward control 35 Case study Nansemond Feedforward control for Limiting aeration FB control more robust, FF requires safety factors against model failures Simple model may not be accurate enough, complex model needs several inputs Increased risk More complex/more expensive Effluent ammonia concentration changes slowly Often not required Very limited control authority at higher ammonia conc. Not functional 36 18
Case study Nansemond Feedforward control for Volume control 37 Case study Nansemond Min/Max blower capacity 12 C 20 C 30 C 38 19
Case study Nansemond Minimum mixing requirements Aeration zone 3, 20 C Min. mixing requirement based on 0.12 scfm/ft 2 (re-suspension) Aeration zone 2, 20 C 39 Case study Nansemond Air flow per diffuser Control scenario 1, 20 C Control scenario 2a, 20 C 40 20
Conclusions I/II Ammonia-based aeration control has two objectives: Limiting aeration to prevent complete nitrification Reduce ammonia effluent peaks Limiting aeration: o energy savings, improved denitrification / bio-p, less carbon addition High control authority to limit nitrification Does NOT increase nitrification capacity when DO > 1.5 mg/l Reducing ammonia effluent peaks: o Ammonia effluent peaks often due to limited mass of nitrifiers Kinetic constraint and cannot be solved by more air Very low control authority of aeration intensity control Swing zones to control ammonia peaks 41 Conclusions II/II Ammonia-based aeration control: What is the control objective? Is FF really necessary? (home-made peaks) Feedforward aeration control often o involves higher risks o is more complex / more expensive o has limited control authority (intensity control) To reduce effluent ammonia peaks, use volume control (swing zones) Use dynamic simulation as a tool to design your process control system! 42 21
Presenter contact information Leiv Rieger Ph.D., P.Eng. inctrl Solutions Inc. Canada Email: rieger@inctrl.ca 43 WEF Webcast Low Energy Process Control Efficient Nutrient Removal under Low Dissolved Oxygen Concentrations Jose Jimenez, Ph.D., P.E. Brown and Caldwell 22
Overview Nitrogen removal What do we know about SND? Factors affecting SND for N removal a. Available carbon b. Dissolved oxygen c. Sludge bulking Applicability and Implications Conclusions 45 Conventional Biological Nutrient Removal N and P removal generally are carried out with physically separated anaerobic, anoxic and aerobic zones N removal relies primarily on autotrophic nitrification and heterotrophic denitrification 23
Simultaneous Nitrification-Denitrification Biological process where nitrification and denitrification occur concurrently in the same aerobic reactor (or in the same floc). Sludge settling characteristics are a real concern SND relies on achieving a dynamic balance between nitrification and denitrification SND depends on: Micro environment Macro environment Bulk DO concentration Carbon availability Presence of novel microorganisms 47 Simultaneous Nitrification-Denitrification Potential Advantages Elimination of separate tanks and internal recycle systems for denitrification Simpler process design Reduction of carbon, oxygen, energy and alkalinity consumption Potential Disadvantages Limited controlled aspects of the process such as: floc sizes internal storage of COD DO profile within the flocs Sludge bulking; primarily because of the excessive growth of filamentous bacteria Complex instrumentation 48 24
Complete Nitrification-Denitrification 49 Complete Nitrification-Denitrification 4.57 mg O 2 / mg ammonia-n nitrified (-) 2.86 mg O 2 /mg N denitrified 1.71 mg O 2 /mg-n removed 50 25
Nitritation-Denitritation Nitritation-Denitritation 3.43 mg O 2 / mg ammonia-n nitrified (-) 1.72 mg O 2 /mg N denitrified 1.71 mg O 2 /mg-n removed 26
Factors affecting SND for N removal Effect of Influent Carbon on SND To accomplish denitrification in any process, the availability of readily biodegradable organic carbon is essential Jimenez et al. (2010) Jimenez et al. (2011) 54 27
Effect of DO on SND Control of bulk DO concentration in the system is essential for achieving a high degree of SND Jimenez et al. (2010) 55 Nitrification fundamentals DO constraints nitrification kinetics 56 28
Effect of DO on Nitrification Nitrification rate at low DO remains at 85% of the maximum value after adaptation Giraldo et al. (2011) 57 New Tools for SND Control Ammonia-based Aeration Control Allows stringent control over DO provided Control aerobic SRT to be as long as needed NOB Repression Rapid transient anoxia seems to be the key Mechanisms? AOB always at maximum growth rate (aerobic SRT control with excess NH4 available) NOB enzyme expression delay Aerobic SRT controlled Nitrite availability delay Oxygen affinity Free ammonia (NH3) inhibition of NOB 29
Low DO Bulking and SND Plant SVI (ml/g) Iron Bridge 115/165 Eastern Reg. 120/160 Snapfinger 200/300 Central 140/180 Winter Haven 130/190 Mandarin 150/180 Marlay Taylor 170/280 Stuart 212/350 Smith Creek 200/245 Lu-Kwang et al. (2006) 59 SND Constant Aeration (Continues Flow) Bulk DO Controlled to 0.5 mg/l 60 30
SND Constant Aeration (Batch Reactor) SND Cyclical Aeration (Continues Flow) 62 31
SND Cyclical Aeration (Batch Reactor) 63 Batch Tests Results Seem to Indicate: AOB rates are not significantly affected during SND NOB seems not to be inhibited by low DO conditions during SND NOB rate slow down during cyclical aeration Possible Nitrite Shunt NO 2 from nitritation can be used for denitritationby Heterotrophs and convert to N 2 Less carbon might be required to convert N to N 2 during SND 64 32
SND - Nitrate vs. Nitrite High C:N Ratio SND (via Nitrate) NH4-based aeration control Lower energy generation potential (WAS-only anaerobic digestion) SND - Nitrate vs. Nitrite PST reduces C:N Ratio; hence, possible C limitations for denitrification SND or Nitrite-Shunt selection based on C:N Ratio NH4-based aeration control Good energy generation potential 33
SND - Nitrate vs. Nitrite Low C:N Ratio for denitrification Nitrite-Shunt required for N removal NH4-based aeration control Nitrite-Shunt compatible with mainstream Anammox High energy generation potential Conclusions (I/III) The application of SND processes may be based and limited by: Influent C:N ratio Sludge bulking issues due to the excessive growth of filamentous bacteria Instrumentation and control requirements The operator has limited control over important parameters impacting SND 68 34
Conclusions (II/III) COD:N ratios of at ~ 8 and ~ 5 are required for SND and Nitrite-Shunt. Optimum bulk DO from 0.2 mg/l to 0.7 mg/l SND is more susceptible to nitrification limitations (DO) and denitrification limitations (carbon). Advantage of cyclical aeration resulted from the more ready availability of NO 2 and NO 3 (generated during nitrification) for denitrification Under constant low DO, denitrification would rely on the slow diffusion of NO 2 and NO 3 from the outer nitrification zone of the flocs into the inner denitrification zone 69 Conclusions (III/III) Nitrite shunt might be possible during SND systems with transient anoxia. The results suggested that the nitrite shunt might take place mainly because of the disrupted nitrification at low DO conditions and pressure to the NOB Cyclical aeration seems to be more effective than constant aeration in avoiding low DO bulking 70 35
Presenter contact information Jose Jimenez Ph.D., P.E. Email: jjimenez@brwncald.com 71 WEF Webcast Low Energy Process Control High Rate Activated Sludge System for Carbon Removal Jose Jimenez, Ph.D., P.E. Brown and Caldwell 36
Overview Evolution of high-rate activated sludge (HRAS) systems Fundamentals and design considerations Solids Retention Time (SRT) Dissolved Oxygen (DO) Case study Strass WWTP, AT 73 Acknowledgement Dr. Charles Bott, HRSD Mark Miller, VT Dr. Sudhir Murthy, DC Water Dr. Bernhard Wett, ARA Consult 7 37
Evolution of HRAS Systems HRAS process uses high F/M ratios and low SRT with short HRT to remove organics from wastewater. Current application of this process recognizes: Particulate and colloidal organics are removed by bio-flocculation (adsorption into the biological floc) and subsequent solids-liquid separation Soluble organics can be removed by intracellular storage, biosynthesis or biological oxidation Chase, ES and Eddy, HP (1944), Sewage Works Journal, Vol. 16, No. 5, pp. 878-885 7 Evolution of HRAS Systems The issue with aerobic treatment is that electrical energy needed for aeration is used to remove chemical energy. This practice is needed by current technology limits for carbon removal in secondary plants. Aerobic treatment is currently the only reliable means to remove carbon to meet secondary limits. 7 38
Evolution of HRAS Systems HRAS systems can be designed and operated as: Carbon oxidation (energy intensive) systems to meet secondary effluent standards. Carbon adsorption processes (less energy intensive) when use as the first step in a two-stage process. Common design parameters: SRT < 3.5 days F:M:0.5-1.0 g BOD per g VSS Detention Time: 0.3 3 hours MLSS: 1,000 to 3,000 mg/l DO: > 2.0 mg/l SRT < 0.5 day F:M: 2.0-10 g BOD per g VSS Detention Time: ~ 0.5 hours MLSS: 1,000 to 3,000 mg/l DO: < 1.0 mg/l (intermittent aeration) 7 A/B Process Alternative HRAS process is operated to minimize the aeration energy needed and to maximize the carbon sorption onto biomass, which is subsequently sent to anaerobic digestion for energy recovery. Figure provided by Dr. Charles Bott, HRSD 7 39
Evolution of HRAS Systems When a HRAS system is the first step in a two-stage process, the picture is substantially different. By operating at low SRT and low DO, carbon oxidation should be minimized and biological flocculation and intracellular storage of soluble substrate (carbon sorption) should be maximized. The transfer of organics from the liquid train to the anaerobic digesters is maximized; hence, energy generation potential can be maximized. 7 Process Control Variables Affecting COD Removal in the HRAS Process 80 40
Impact of the System SRT on WAS VS Content 81 Impact of SRT on the C Removal Efficiency 82 41
Impact of DO on the COD Removal Efficiency 83 Impact of SRT on the Specific Aeration Requirement At Lower SRT, the SAR decreases indicating possible C adsorption and storage. 84 42
Process Control Variables Affecting COD Removal in the HRAS Process 85 HRSD s A-Stage Pilot Plant High CO 2 PR = High OHO activity which may indicate C oxidation (energy intensive process) Low CO 2 PR = Lower OHO activity which may indicate C adsorption and storage (less energy intensive process) Figure provided by Mark Miller, VT/ HRSD 86 43
Case Study - Strass WWTP, AT Two-stage BNR plant (A/B plant) Load variations from 90,000 to 230,000 PE weekly average Data provided by Dr. Wett 87 Case Study - Strass WWTP, AT Figure provided by Dr. Wett 88 44
A-Stage: 0.5 days SRT 55-65% COD removal B-Stage: 10 days SRT Pre-denitrification, on-line NH4-N controlled intermittent aeration Brown Data provided and Caldwell by Dr. Wett 89 Strass WWTP - High Gas Potential in A-Stage Sludge Compared to B- Stage Sludge Data provided by Dr. Wett 90 45
Strass WWTP - Maximize Transfer of Organics from Liquid Train to the Digesters Means Operation at Low SRT or High F/M Ratio Data provided by Dr. Wett 91 Strass WWTP - Multi-Step Optimization Process both in Energy Consumption and Production Data provided by Dr. Wett 92 46
Conclusions Carbon oxidation = energy intensive system. Carbon adsorption processes = less energy intensive system. The proper selection of SRT (F:M), HRT and DO, bioflocculation and intracellular storage of carbon should be maximized. The transfer of organics from the liquid train to the anaerobic digesters is maximized; hence, energy generation potential can be maximized. Presenter contact information Jose Jimenez Ph.D., P.E. Email: jjimenez@brwncald.com 94 47
Questions? 48