Model predictive control of a fermentation feed in biofuel production

Abstract

System and method for managing fermentation feed in a biofuel production process, comprising a dynamic multivariate predictive model-based controller coupled to a dynamic multivariate predictive model. The model is executable to: receive process information, including water inventory and biomass information, from the biofuel production process; receive a specified objective for the fermentation feed specifying a target biomass concentration; and generate model output comprising target values for a plurality of manipulated variables of the biofuel production process, including target flow rates of water and/or biomass contributing to the fermentation feed in accordance with the specified objective. The controller is operable to dynamically control the biofuel production process by adjusting the plurality of manipulated variables to model-determined target values to stabilize water/biomass balance in the fermentation feed in accordance with the specified objective, including the specified target biomass concentration.

1 - 37 . (canceled) 
     
     
         38 . A computer-implemented method for managing fermentation feed in a biofuel production process, comprising:
 providing a dynamic multivariate predictive model of biomass concentration and water inventory for a fermentation feed of the biofuel production process;   receiving a specified objective for the fermentation feed specifying a target biomass concentration for the fermentation feed;   receiving process information, comprising water inventory information and biomass concentration information, from the biofuel production process;   executing the dynamic multivariate predictive model in accordance with the objective using the received water inventory and biomass concentration information as input, thereby generating model output comprising target values for a plurality of manipulated variables of the biofuel production process, including target water and biomass flow rates contributing to the fermentation feed in accordance with the objective; and   controlling the biofuel production process, including water and biomass flow rates related to the fermentation feed of the biofuel production process, in accordance with the plurality of manipulated variables to stabilize water/biomass balance in the fermentation feed in accordance with the specified objective, including the specified target biomass concentration.   
     
     
         39 . The method of  claim 38 , further comprising:
 receiving constraint information specifying one or more constraints; and   executing the dynamic multivariate predictive model in accordance with the objective using the received water inventory, the one or more constraints, and biomass concentration information as input, thereby generating the model output in accordance with the objective and subject to the one or more constraints.   
     
     
         40 . The method of  claim 39 , wherein the one or more constraints comprise one or more of:
 mill amperage limit that limits fermentation feed rate or cook flow rate;   thin stillage tank level limits that limit fermentation residual broth recycle rate;   slurry water tank level limits that limit water addition rates;   pressure limits that limit pump capabilities, safety or processing rates;   back-end processing rate limits that limit fermentation fill rate; or   cook temperature limits that limit fermentation fill rate;   wherein the objective further comprises one or more sub-objectives comprising one or more of:
 a target fermentation feed rate; 
 a consistent fermentation residual broth fraction of the fermentation feed; or 
 a consistent water fraction of the fermentation feed; 
   wherein the dynamic multivariate predictive model comprises a multivariate predictive model that represents relationships between the one or more constraints, the objective, including the one or more sub-objectives, and the plurality of manipulated variables.   
     
     
         41 . The method of  claim 39 , wherein one or more constraints comprise one or more of:
 water constraints;   biomass constraints;   feed constraints;   equipment constraints;   capacity constraints;   temperature constraints;   pressure constraints;   energy constraints;   market constraints;   economic constraints; or   operator imposed constraints.   
     
     
         42 . The method of  claim 39 , wherein the one or more constraints includes equipment constraints comprising one or more of:
 operating limits for pumps;   operational status of pumps;   tank capacities;   operating limits for tank pressures;   operational status of tanks;   operating limits for valve pressures;   operating limits for valve temperatures; or   operating limits for pipe pressures.   
     
     
         43 . The method of  claim 38 , wherein the specified objective comprises one or more of:
 one or more operator specified objectives;   one or more predictive model specified objectives;   one or more programmable objectives;   a target feed rate for the fermentation feed;   one or more cost objectives;   one or more quality objectives;   one or more equipment maintenance objectives;   one or more equipment repair objectives;   one or more equipment replacement objectives;   one or more economic objectives;   a target throughput for the biofuel production process;   one or more objectives in response to emergency occurrences;   one or more dynamic changes in water inventory information;   one or more dynamic changes in biomass concentration information; or   one or more dynamic changes in one or more constraints on the biofuel production process.   
     
     
         44 . The method of  claim 38 , wherein controlling the flow rates of water inventory and biomass concentration comprises:
 operating one or more flow controllers coupled to the dynamic multivariate predictive model, wherein the one or more flow controllers control one or more of:
 mill speed; 
 pump speeds for one or more biomass feeds; or 
 pump speeds for one or more water feeds. 
   
     
     
         45 . The method of  claim 38 , wherein the water inventory information includes one or more of:
 fluid levels for one or more water tanks;   capacity limits for each of the one or more water tanks;   operational status for each of the one or more water tanks; or   flow rates for one or more water flows.   
     
     
         46 . The method of  claim 45 , wherein the one or more water flows comprise one or more of:
 water flow rates to each of one or more processing units;   recycled water flow rates from one or more distillation units; or   recycled water flow rates from one or more stillage processing units.   
     
     
         47 . The method of  claim 38 , wherein the water inventory information comprises inventory information for a plurality of water sources, comprising one or more of:
 one or more water sources; or   one or more recycled water sources; and   wherein the target water flow rates comprise target flow rates for one or more of:   water; or   recycled water.   
     
     
         48 . The method of  claim 47 , wherein the recycled water comprises one or more of:
 fermentation broth recycle water;   evaporator condensate recycle water;   distillation bottoms recycle water; or   treated water from an anaerobic digester.   
     
     
         49 . The method of  claim 47 , wherein executing the dynamic multivariate predictive model comprises the dynamic multivariate predictive model:
 determining a total water flow rate for the fermentation feed in accordance with the objective; and   partitioning the total water flow rate for the fermentation feed among the plurality of water sources to determine respective target flow rates for each of the plurality of water sources; and   wherein controlling the flow rates of water inventory comprises controlling respective flow rates for each of the plurality of water sources in accordance with the respective target flow rates.   
     
     
         50 . The method of  claim 49 , wherein partitioning the total water flow rate for the fermentation feed among the plurality of water sources to determine respective target flow rates for each of the plurality of water sources comprises performing the partitioning subject to one or more constraints regarding the plurality of water sources. 
     
     
         51 . The method of  claim 50 , wherein the one or more constraints regarding the plurality of water sources comprise one or more of:
 tankage constraints, comprising one or more of:
 limits on tank levels; 
 limits on tank fill or emptying rates; 
   impurity constraints, comprising one or more of:
 percent impurity constraints; or 
 constraints on impurity type; or 
   constraints on water transport.   
     
     
         52 . The method of  claim 51 , wherein the constraints on water transport comprise one or more of:
 operational status of one or more water transport elements; or   operational limits on one or more water transport elements.   
     
     
         53 . The method of  claim 38 , wherein the biomass concentration information comprises one or more of: feed rates for each mill, amp limits for each mill, water flow rates, stream density, temperature or biomass concentration limit. 
     
     
         54 . The method of  claim 38 , wherein the dynamic multivariate predictive model specifies relationships between fermentation feed rates and equipment constraints, the method further comprising:
 wherein the constraint information comprises one or more equipment constraints of the biofuel production process;   wherein the specified objective comprises a target feed rate for the fermentation feed; and   wherein executing the dynamic multivariate predictive model comprises executing the dynamic multivariate predictive model using the received water inventory and biomass concentration information, and the one or more equipment constraints as input, and wherein the target water and biomass flow rates are computed to approach and maintain the target feed rate for the fermentation feed subject to the one or more equipment constraints.   
     
     
         55 . The method of  claim 38 , wherein the dynamic multivariate predictive model further specifies relationships between biomass fractional flow rates and enzyme flow rates, the method further comprising:
 receiving enzyme flow rate information from the biofuel production process comprising one or more enzyme flow rates;   wherein executing the dynamic multivariate predictive model comprises executing the dynamic multivariate predictive model using the received water inventory and biomass concentration information, constraint information, and the enzyme flow rates as input, and wherein the model output further comprises target enzyme flow rates; and   wherein controlling the flow rates of water inventory and biomass concentration further comprises controlling enzyme flow rates related to the fermentation feed of the biofuel production process in accordance with the target enzyme flow rates to stabilize water/biomass and enzyme/biomass balance in the fermentation feed.   
     
     
         56 . The method of  claim 55 , wherein the enzyme flow rates comprise respective ratios of enzyme addition rates to biomass addition rates. 
     
     
         57 . The method of  claim 55 , wherein controlling the enzyme flow rates in accordance with the target enzyme flow rates is performed to maintain the fermentation feed at a specified target feed rate. 
     
     
         58 . The method of  claim 57 , wherein the dynamic multivariate predictive model further specifies at least one equipment setting as a capacity constraint, and wherein controlling the enzyme flow rates comprises controlling enzyme ratios to increase milling and cook capacity in the biofuel production process by increasing liquefaction agents to reduce slurry viscosity and allow higher flow rates within limits due to the at least one equipment setting. 
     
     
         59 . The method of  claim 58 , wherein the at least one equipment setting comprises at least one valve position, and wherein the limits due to the at least one equipment setting comprise line pressure limits or control valve ranges. 
     
     
         60 . The method of  claim 38 , wherein the biomass concentration information is provided by another model with inputs comprising one or more of:
 measured temperature of the biomass;   mill output;   mill energy use; or   water input to the fermentation feed.   
     
     
         61 . The method of  claim 60 , wherein at least a portion of the water input to the fermentation feed is determined by indirect measurements related to the water input. 
     
     
         62 . The method of  claim 60 , wherein the other model is a dynamic multivariate predictive model that includes time-response of biomass density to changes in one or more inputs to the biofuel production process. 
     
     
         63 . The method of  claim 60 , wherein the one or more inputs to the biofuel production process are dynamically filtered based on the time-response and provided as inputs to the other model, and wherein one or more controllers for controlling biomass concentration are configured in accordance with the time-response. 
     
     
         64 . The method of  claim 38 , wherein the dynamic multivariate predictive model specifies one or more temperature control constraints for the biofuel production process, the method further comprising:
 receiving one or more temperature targets and a slurry throughput target for the fermentation feed;   receiving temperature information from the biofuel production process;   wherein executing the dynamic multivariate predictive model comprises executing the dynamic multivariate predictive model using the received water inventory and biomass concentration information, and the temperature information as input, and subject to the one or more temperature control constraints, wherein the model output further comprises one or more target temperature values, and wherein, in computing the model outputs, the dynamic multivariate predictive model balances errors in the one or more temperature targets vs. the slurry throughput target; and   wherein controlling the flow rates of water inventory and biomass concentration further comprises controlling temperature related to the fermentation feed of the biofuel production process in accordance with the target temperature values to stabilize water/biomass and enzyme/biomass balance in the fermentation feed.   
     
     
         65 . The method of  claim 64 , wherein the temperature information from the biofuel production process comprises one or more of:
 cook temperature;   hydroheater temperature;   cook flash temperature; or   liquefaction temperature; and   wherein the one or more target temperature values comprise one or more of:
 a target cook temperature; 
 a target hydroheater temperature; 
 a target cook flash temperature; or 
 a target liquefaction temperature. 
   
     
     
         66 . The method of  claim 38 , wherein the dynamic multivariate predictive model specifies pH relationships in the biofuel production process, the method further comprising:
 receiving pH information from the biofuel production process;   wherein executing the dynamic multivariate predictive model comprises executing the dynamic multivariate predictive model using the received water inventory and biomass concentration information, and the pH information as input, and wherein the model output further comprises a target pH for the fermentation feed, and a pH control agent target; and   wherein controlling the flow rates of water inventory and biomass concentration further comprises controlling pH related to the fermentation feed of the biofuel production process in accordance with the target pH and pH control agent target.   
     
     
         67 . The method of  claim 66 , wherein the pH relations in the biofuel production process comprise relations between pH and one or more of:
 mixing tank level of a pH control agent with slurry;   ambient temperature;   time of day;   solids content of the fermentation feed;   liquid temperature of the fermentation feed; or   slurry flow rate of the fermentation feed;   wherein the pH information comprises one or more of:
 the mixing tank level of the pH control agent with slurry; 
 the ambient temperature; 
 the time of day; 
 the solids content of the fermentation feed; 
 the liquid temperature of the fermentation feed; or 
 the slurry flow rate of the fermentation feed. 
   
     
     
         68 . The method of  claim 38 , wherein executing the dynamic multivariate predictive model in accordance with the objective comprises:
 an optimizer executing the dynamic multivariate predictive model in an iterative manner to determine values of the plurality of manipulated variables that satisfy the objective subject to one or more constraints to determine the target values for the manipulated variables.