combined heat & power (chp)utw10249.utweb.utexas.edu/.../projectspresentation.pdf · combined...
TRANSCRIPT
Combined Heat & Power (CHP) Research Interests: Modeling, control, and
optimization of CHP systems in a smart grid environment
Incorporation of thermal storage to improve the flexibility of CHP systems
Wesley Cole
This work is done in collaboration with the smart grid demonstration project carried out by Pecan Street Project, Inc.
Tumor Irrigation Model for Virotherapy Cancer Treatment Ricardo Dunia(UT Austin)
Achievements: • Analyze existent dynamic models and equilibrium conditions for existent models. • Propose a new model type to account for the effect of blood irrigation in tumors undergoing Virotherapy. Future Work: - Make the irrigation model suitable for model predictive control.
Figure 2- Irrigation model Figure 1- Cancer Growing/Shirnking under different dose.
Z(t)
P(t)Host
Controller
Tumor
Multistate PCA and Multistate PLS for Continuous Processes Ricardo Dunia(UT Austin)
Achievements: • Novel methodology to define operating regions and transition trajectories between states of operation. • These techniques enhance the procedure to normalize process variables for process monitoring and fault detection. Future Work: -Online implementation for real-time applications
Figure 2- Transition trajectories between operating points Figure 1- State Variable used for CO2
Capture process
Networked Control with Wireless Field Devices
Douglas French (UT Austin) • Achievements:
1) Demonstrated missing data has a minimal effect on control performance for practical data loss rates. 2) Incorporation of battery lifetime models into control system design. 3) Demonstrated state estimation can be used to extend the time the system runs open loop to extend battery life
1) Develop performance index incorporating battery usage 2) Investigate effects of different event triggered sampling on battery life and control performance 3) Develop optimal tuning guidelines
• Future plans:
Control Performance Assessment (CPA) Xiaojing Jiang, Ph.D. Dec 2012
Sponsored by NSF/GOALI • Achievements:
Developed a new method to determine potential improved performance and improved controller tuning factors;
Developed a new CPA method to determine whether the control is optimal or not (treatment of time delay and high-mix parameters). See Figure below for control chart range.
• Future plans:
New Traditional
Apply the new method to DMOS6
Extend the new method to other R2R controllers
TI Sponsor: John Stuber
0 200 400 600 800 1000
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0 200 400 600 800 1000
-0.2
0.0
0.2
0.4
Optimal SuboptimalOptimal
PI-H
arris
' typ
e
Run number
Suboptimal
PI-n
ew m
etho
dRun number
Optimal
Suboptimal
Suboptimal
Energy and Environmental Efficiency in Semiconductor Manufacturing
Kriti Kapoor, Ph.D. Aug 2013
o Achievements • Developed an optimizer around CleanCalc II to minimize energy use
– When air handler exit air temperatures (cooling and heating mode) and dew point are varied, the program converges at 66.2, 59.7 and 49.7 F respectively; which represents annual savings of ~600,000 kWh from the initial point of operation.
• Modified the Lagrangian solution to improve precision
o Future work • Modeling of greenhouse gases (GHG)
emissions from semiconductor factory • Analysis of scope of heat integration
in fabs
o TI Sponsor: Sunil Thekkepat
Improvements of the Capacitance Resistive Model (CRM)
1) Application of the CRM on Synfield with some data missing 2) Nonlinear CRM which depends on initial production rate only 4) Linear CRMP (ICR model) using cumulative water injected and cumulative total
liquid produced 5) Confidence interval test via the ICR model 6) Recursive Least Squares (RLS) parameters estimation to take into account dual role
wells and wells shut-in 7) Introduction of space-time variant model parameters in the CRM or the ICR model 8) Modeling gas reservoirs using pseudo-pressure
Jongsuk Kim (UT Austin)
• Achievements:
• Future plans:
1) Application of the ICR model on real field 2) Extending the models for use in - gas floods (CO2), chemical floods, and
thermal recovery 3) Application of the ICR model on horizontal
or inclined wells
Figure 1. Schematic representation of the impact of an injection rate signal on total production response for an arbitrary reservoir control volume in the CRM.
Multi-State PLS for Continuous Process Analytics Vinay Kumar (UT Austin)
Achievements:
- Analyzed various data driven techniques being used in predictive analytics in process control. - Simulations used to implement a novel PLS/PCA based process monitoring technique.
- Developed a MATLAB based GUI for the implementation of Process Analytics for steady state. Future Work: - Field Test of the Prototype developed - Modification of GUI to incorporate System Dynamics.
Figure 2- Predicted and Actual response for simulation using Multi-state PLS
Figure 1- Simulation used to implement Multi-state PLS using GUI.
Enhancement of Capacitance Resistive Model (CRM) for Primary, Secondary and Tertiary Recovery
• Achievements: 1. For secondary recovery application: Evaluate consistency of CRM results, calculate sensitivity
of CRM results on radial limit, analyze fitting window selection, study CRM forecast capability, apply CRM optimization on part of the field and analyze the results
2. Propose new model: integrated capacitance-resistive (ICR) to determine producer productivity index and dynamic pore volume and average reservoir pressure for oil primary production, validate ICR on various synthetic fields and actual field. data
3. Develop ICR model for primary gas reservoir, validate on synthetic fields and actual field data 4. Study CRM applicability on water alternating gas (WAG) injection field data with different
weight on water injection rate
Anh Nguyen
• Future plans: 1. Study data noise and co-linearity effects on
CRM results 2. Develop CRM model for WAG field and
validate on synthetic data
CRM uses tank model concept to infer reservoir characteristics. This research is to develop CRM to apply the model on real oil fields.
Fig. 1: Estimated dynamic pore volumes of wells P1 and P2 in synthetic primary recovery field vs. field actual pore volume
5. Develop and validate the method to infer transmissibility from CRM results
6. Compare CRM results to tracer test data and synthetic field streamline results
Patient-Specific Insulin Therapy Ramiro Palma (UT Austin)
• Project Summary – Model identification and simulations to support insulin
therapy regimens in T1DM and T2DM
• Work to date – Large-scale in silico epidemiological basal insulin therapy in
T2DM – Development of physiological model for interstitial glucose
dynamics – System identification using real data in T1DM
Modeling, Optimization and Control Systems for Thermal Energy Storage Systems
Kody Powell, Ph.D. Aug 2013 (Sponsored by NSF)
Thermal Energy Storage can be used with Concentrated Solar Power Plants to level electricity generation rates and extend generation
periods to when the sun is night shining.
Thermal Energy Storage can be used for cooling applications. Ice is made at night, which is then used during the day to cool buildings.
Research Goals: •Develop models and base control schemes for:
•Solar thermal systems •Energy storage for peak shifting •Combined heat and power
•Use weather forecasts and predicted consumer behavior to optimize the performance of the energy system dynamically
People and Power: Modeling Demand Response
Research Goals • To model and optimize peak energy
consumption at the residential level • Focus on the impact of pricing signals,
photovoltaics, energy storage, and flexible loads (electric vehicles)
Highlights • Matlab/Simulink modeling platform • Collaboration with Pecan Street Inc.
(smart grid demonstration project) • Opportunity to utilize real data
from 1000+ homes • Interoperability of models and
collaboration across UT and industry
Demand Response- Electricity price reduction as an output of decreased load Figure Credit: Pecan Street Project
Current generation home energy controller, this one produced by Cisco Systems, Inc.
Akshay Sriprasad
Dynamic Modeling, Optimization and Control of Amine Plant to Remove CO2 from Coal Fired Power Plants
Sepideh Ziaii , Dr. Gary Rochelle , Dr. Thomas Edgar
Luminant Carbon Management Program Industrial Associates Program for CO2 Capture by Aqueous Absorption
Absorber
Stripper
LP IP HP
LT LC LT LC
Flue gas from power plant
Modeling in Aspen Custom Modeler®: •Dynamic rate-based model of packed columns (absorber & stripper) and reboiler • Simplified steady state model of heat exchangers •General performance curve for compressor and pumps •Ellipse law for steam turbines
Optimization and control : • Multi-variable optimization : Optimize manipulating variables to minimize energy in response to partial load operational scenarios • Explore optimum control strategies for dynamic operations •Investigate the effects of hold up in the storage tanks •Application of variable speed compressor in operation