Wastewater treatment plant online monitoring and control

Abstract

A method of operating a waste water treatment plant (WWTP) having at least one of an aerobic digester (AD) and a membrane bioreactor (MBR) is described. The method of operating AD is comprised of monitoring and controlling AD in real-time using an online extended Kalman filter (EKF) having a online dynamic model of AD. The EKF uses real-time AD measured data, and online dynamic model of AD to update adapted model parameters and estimate model based inferred variables for AD, which are used for AD control by AD control system having supervisory and low-level control layers. The method of operating MBR is similar to that of AD. The supervisory control ensures the WWTP satisfying the effluent quality requirement while minimize the operation cost. A WWTP having at least one of an AD or MBR is disclosed. The method of operating a WWTP can be implemented using a computer.

What is claimed is: 
     
         1 . A method of monitoring and controlling the operating conditions of a membrane bioreactor (MBR), comprising:
 providing a MBR;   monitoring said MBR, wherein said monitoring comprises:
 providing a MBR offline extended Kalman filter (EKF) having an offline dynamic model of said MBR, providing a MBR online EKF having an online dynamic model of said MBR; wherein said offline and said online dynamic models of said MBR are comprised of states, process material balances, energy balances, bio-chemical reaction kinetics, estimated parameters, and adapted model parameters; wherein said adapted model parameters are a subset of said estimated parameters; 
 providing historical operation data for said MBR, wherein said historical operation data is comprised of historical measured input data, historical measured output data, and historical laboratory analysis data; 
 identifying said estimated parameters of said offline dynamic model of said MBR using said MBR offline EKF and said historical operation data for said MBR; 
 importing said estimated parameters from said offline dynamic model of said MBR into said online dynamic model of said MBR; 
 providing real time operation data for said MBR to said MBR online EKF, wherein said real time operation data is comprised of real time measured input data and real time measured output data of said MBR; 
 updating said adapted model parameters of said online dynamic model of said MBR and estimating one or more model based inferred variables of said MBR using said MBR online EKF, said online dynamic model of said MBR, said real time measured input data of said MBR, and said real time measured output data of said MBR; and 
 providing one or more of said adapted model parameters of said online dynamic model of said MBR and said model based inferred variables of said MBR to an operator of said MBR; 
 wherein limits are applied to one or more of said estimated parameters and said adapted model parameters; wherein constraints are applied to one or more of said model based inferred variables; 
   controlling said MBR, wherein said controlling comprises:
 providing an MBR control system; 
 wherein said MBR is comprised of an aerobic tank, a membrane tank, and optionally an anoxic tank; 
 wherein said MBR control system uses one or more of said real time measured input data of said MBR, said real time measured output data of said MBR, said estimated parameters of said online dynamic model of said MBR, or said model based inferred variables of said MBR to control at least one of a pH of said anoxic tank, a pH of said aerobic tank, fluid level of said aerobic tank, dissolved oxygen (DO) concentration of said aerobic tank, mixed liquor suspended solids (MLSS) concentration of said membrane tank, biodegradable COD (bCOD) addition flow rate setpoint of said anoxic tank, at least one nutritional additive concentration of said anoxic tank, or at least one recycle flow setpoint of said MBR; 
 wherein said MBR control system is comprised of an MBR supervisory control system and an MBR low-level control system. 
   
     
     
         2 . The method of  claim 1 , wherein said MBR is comprised of an aerobic tank, a membrane tank, and optionally an anoxic tank; wherein said aerobic tank is located upstream of said membrane tank; wherein said anoxic tank is located either immediately upstream or downstream of said aerobic tank when said anoxic tank is present. 
     
     
         3 . The method of  claim 2 , wherein said aerobic tank and said anoxic tank are modeled separately in both of said online and offline dynamic models of said MBR when both of said aerobic and said anoxic tanks are present. 
     
     
         4 . The method of  claim 2 , wherein said MBR is further comprised of a mixer and at least one recycle line. 
     
     
         5 . The method of  claim 4 , wherein said at least one recycle line of said MBR is an anoxic tank recycle line or an aerobic tank recycle line. 
     
     
         6 . The method of  claim 1 , wherein materials for said material balances in said online and offline dynamic models of said MBR are comprised of at least one of particulate inert, slowly degradable substrate, heterotrophic biomass, autotrophic biomass, decayed biomass, soluble inert, soluble readily degradable substrate, dissolved oxygen, dissolved nitrate-N(Nitrogen), dissolved ammonia-N, particulate bio-degradable-N, or bicarbonate alkalinity. 
     
     
         7 . The method of  claim 1 , wherein said bio-chemical reaction kinetics in said online and offline dynamic models of said MBR are comprised of at least one of aerobic heterotroph, anoxic heterotroph, aerobic autotroph, decay of heterotroph, decay of autotroph, ammonification of soluble organic N, hydrolysis of organics, or hydrolysis of organic N. 
     
     
         8 . The method of  claim 1 , wherein said historical operation data of said MBR and said real time operation data of said MBR are comprised of at least one of raw influent pH, raw influent temperature, raw influent flow rate, raw influent total organic carbon (TOC), raw influent total inorganic carbon (TIC), added alkali flow rate, added alkali concentration, effluent flow out rate, raw influent soluble chemical oxygen demand (SCOD), raw influent total chemical oxygen demand (TCOD), raw influent readily biodegradable chemical oxygen demand (COD), raw influent slowly biodegradable COD, raw influent volatile suspended solids (VSS), raw influent total suspended solids (TSS), raw influent nitrate nitrogen, raw influent ammonia-nitrogen, raw influent soluble biodegradable organic nitrogen, raw influent particulate degradable organic nitrogen, raw influent inorganic inert particulate, membrane permeate flow rate, wasting sludge flow rate, anoxic tank addition biodegradable COD flow, anoxic rank reactor pH, anoxic tank Dissolved Oxygen, anoxic tank temperature, anoxic tank liquid level, anoxic tank mixed liquor volatile suspended solids (MLVSS), anoxic tank MLSS, aerobic rank blower air flow rate, aerobic tank reactor pH, aerobic tank alkalinity, aerobic tank MLVSS, aerobic tank MLSS, aerobic tank Dissolved Oxygen, aerobic tank temperature, aerobic tank liquid level, membrane tank MLSS, membrane tank MLVSS, membrane permeate SCOD, membrane permeate TCOD, membrane permeate TOC, membrane permeate TIC, membrane permeate nitrate nitrogen, membrane permeate ammonia-nitrogen, wasting sludge MLSS, or wasting sludge MLVSS. 
     
     
         9 . The method of  claim 1 , wherein said estimated parameters and said adapted model parameters of said offline dynamic model of said MBR and said online dynamic model of said MBR are comprised of at least one of hetrotrophic maximum specific growth rate, anoxic/aerobic hetrotroph growth rate, anoxic/aerobic hydrolysis rate fraction, particulate hydrolysis max specific rate constant, autotrophic maximum specific growth rate, decay constant for heterotrophs, decay constant for autotrophs, yield of hetrotrophic biomass, yield of autotrophic biomass, carbon content in soluble substrate, carbon content of particulate substrate, carbon content of soluble inert, carbon content of particulate nondegradable organic, mass transfer coefficient for O 2  removal in aerobic tank, or mass transfer coefficient for CO 2  removal in anoxic tank. 
     
     
         10 . The method of  claim 1 , wherein at least one of said estimated parameters of said offline dynamic model of said MBR and said model based inferred variables of said online dynamic model of said MBR are estimated with confidence intervals. 
     
     
         11 . The method of  claim 1 , wherein said model based inferred variables of said online dynamic model of said MBR are comprised of at least one of the following unmeasured inputs or outputs of said MBR: raw influent alkalinity, raw influent nitrate nitrogen, raw influent ammonia-nitrogen, raw influent SCOD, raw influent TCOD, raw influent readily biodegradable COD, raw influent slowly biodegradable COD, raw influent VSS, raw influent TSS, raw influent inorganic inert particulate, anoxic rank SCOD, anoxic tank MLVSS, anoxic tank nitrate nitrogen, anoxic tank ammonia-nitrogen, anoxic tank biodegradable COD, aerobic tank SOCD, aerobic tank MLVSS, aerobic tank nitrate nitrogen, aerobic tank ammonia-nitrogen, aerobic tank biodegradable COD, membrane tank MLVSS, membrane permeate SCOD, membrane permeate biodegradable COD, membrane permeate TCOD, membrane permeate nitrate nitrogen, membrane permeate ammonia-nitrogen, wasting sludge MLVSS, COD removal rate, or nitrogen removal rate. 
     
     
         12 . The method of  claim 1 , further comprising tuning said adapted model parameters of said online dynamic model of said MBR using different weights for said real time operation data and a prior knowledge of measurement accuracy of said real time operation data. 
     
     
         13 . The method of  claim 1 , further comprising adjusting said adapted model parameters of said online dynamic model of said MBR by one or both of:
 calculating model predicted outputs of said MBR using said MBR online EKF, said online dynamic model of said MBR, said real time measured input data of said MBR, and said real time measured output data of said MBR, comparing said measured output data of said MBR and said model predicted outputs of said MBR, and updating said adapted model parameters of said online dynamic model of said MBR such that said real time measured output data of said MBR substantially correspond with said model predicted outputs of said MBR; or   periodically re-identifying said estimated parameters of said offline dynamic model of said MBR using said MBR offline EKF and said historical operation data for said MBR, and importing said estimated parameters from said offline dynamic model of said MBR into said online dynamic model of said MBR.   
     
     
         14 . The method of  claim 1 , wherein at least one of said monitoring said MBR or said controlling said MBR is performed using a computer. 
     
     
         15 . The method of  claim 1 , wherein controlling at least one nutritional additive concentration of said anoxic tank prevents biomass overfeeding and starvation, wherein controlling said pH of said anoxic tank minimizes alkali dosing, wherein controlling said pH of said aerobic tank minimizes alkali dosing, wherein controlling the fluid level of said aerobic tank minimizes the affect of fluid perturbations of said aerobic tank, wherein controlling the dissolved oxygen (DO) concentration of said aerobic tank ensures that a proper concentration of DO is present in said aerobic tank, wherein controlling the MLSS concentration of said membrane tank maximizes membrane permeability, wherein controlling the bCOD addition flow rate setpoint of said anoxic tank minimizes bCOD usage, wherein controlling at least one recycle flow setpoint of said MBR helps to maintain flow through said MBR. 
     
     
         16 . The method of  claim 1 , wherein said MBR supervisory control system is comprised of at least one of an aerobic tank DO supervisory controller, an anoxic tank recycle flow supervisory controller, or an anoxic tank bCOD addition flow rate supervisory control scheme. 
     
     
         17 . The method of  claim 16 , wherein said anoxic tank bCOD addition flow supervisory control scheme of said MBR is comprised of an anoxic tank bCOD setpoint supervisory controller, an anoxic tank bCOD addition flow rate supervisory feedback controller, and an anoxic tank bCOD addition flow rate supervisory feedforward controller. 
     
     
         18 . The method of  claim 1 , wherein said MBR low-level control system is comprised of at least one of an aerobic tank fluid level PI controller, an aerobic tank pH controller, an anoxic tank pH controller, an anoxic tank recycle line flow rate controller, an aerobic tank DO concentration controller, an anoxic tank nutritional additive concentration controller, an aerobic tank recycle line flow rate Proportion-Integration (PI) controller, a total MBR recycle flow rate PI controller, or a membrane tank MLSS concentration controller. 
     
     
         19 . The method of  claim 18 , wherein said membrane tank MLSS concentration controller uses one or more of said model based inferred variables of said MBR. 
     
     
         20 . The method of  claim 19 , wherein said model based inferred variables of said MBR include MLVSS concentration and/or MLSS concentration. 
     
     
         21 . The method of  claim 16 , wherein said aerobic tank DO supervisory controller, said anoxic tank recycle flow supervisory controller, and said anoxic tank bCOD addition flow rate supervisory control scheme satisfy membrane permeate requirements on COD, nitrate, and ammonia, while minimizing aeration, recycle flow, and bCOD addition. 
     
     
         22 . The method of  claim 16 , wherein at least one of said aerobic tank DO supervisory controller, said anoxic tank recycle flow supervisory controller, or said anoxic tank bCOD addition flow rate supervisory control scheme uses at least one of said estimated parameters of said online dynamic model of said MBR or said model based inferred variables of said MBR.