Fall

1
Fall 08 Switchgear Room Analysis and Expectations

• Interdisciplinary Project with South Texas
  Project Funded by
• The Nuclear Power Institute

2
The Team

• Team Members
• Matt Langston CHEN Senior
• Kyle Bowzer MEEN Junior
• Ryan Bigelow MEEN Junior
• Matthew King MEEN Junior
• Richard Colunga ELEN Sophomore
• Jennifer Banegas CVEN Freshman
• George Campa CHEN Freshman
• Brent Mayorga AERO Freshman
• Mentors
• Graduate Mentor Andron Creary
• TAMU Mentor Mr. Cable Kurwitz
• STP Mentor Mr Rick Grantom

3
Agenda

• Motivation Project Components
• Project Objectives
• CFD Analysis
• Lumped Parameter Simulations
• Experimental Results
• Summary
• Future Work

4
Motivation Project Components

• Motivation
• Provide STPs Probabilistic Risk Assessment (PRA)
  Team with air temperature profile data after a
  hypothetical loss of HVAC.
• Project Components
• CFD Analysis
• Using a SolidWorks created model of room and
  electrical cabinets
• Analytical Calculations
• Perform calculations in Matlab using a Lumped
  Parameter Method
• Laboratory Experiments
• Run experiments investigating the heat transfer
  and energy storage within a solid material

5
Simulation Objectives

• Determine air and metal heat up rates during
  various HVAC failure scenarios
• Gain information on when and where the air
  temperature reaches manufacturers critical
  temperature (104F)
• Investigate the effect of energy storage within
  metal in the transformers using a cabinet CFD
  model

6
CFD - SolidWorks Model
Outlets
Inlets
Inlets
Heater Rods
7
Distribution of Heat Loss
29935 watts
12596 watts
1966 watts
200 watts
29935 watts
2234 watts
3131 watts
2519 watts
1759 watts
1523 watts
8
CFD Simulations

• Computer Simulations
• Case 1 Steady State
• Simulates the EAB rooms Normal Operating
  Conditions (50F inlet air temp and 19870 cfm)
• Case 2 Transient
• Simulates the loss of one of the HVAC trains (50
  air flow)
• Case 3 Transient
• Simulates the HVAC chiller failure (73F inlet
  air temp instead of 50F)
• Case 4 Transient
• Simulates the total loss of HVAC

9
Case 1 - Temperature Profile for Normal
Operating Conditions
64F average
10
Case 1 - Maximum Temperature Normal Operating
Conditions
Max air temp 78 F (above main cabinet)
Max air temp between cabinets 64F (at 5ft)
5 ft
11
Case 2 - Temperature Profile Half-flow
Simulation
68F average
SS after 21min
12
Case 3 - Temperature Profile HVAC Chiller
Failure
83F average
SS after 19 min
13
Temperature Profile Total HVAC Failure When
Critical Temperature (104F) is Reached
Critical temperature (104F) location
After 19 minutes
14
Final Results Plot
104F
Case 4
Case 3
Case 2
Case 1
15
Energy Storage in Transformers

• The HVAC failure problem is more complicated
  because it is a transient problem
• Stored thermal energy flow is important in the
  temperature history
• In particular, the heat up of the transformers
  copper windings and iron cores due to the high
  specific heat capacity.
• Bounding the Specific Heat
• Based on manufacturers specifications of
  transformer cabinets in the EAB room, metal mass
  composition values were assumed
• Stainless steel 15-20
• Aluminum 5-20
• Iron 20-60
• Copper 40-60

16
Bounded Values

• Using Matlab, all possible mass combinations were
  computed
• Used to determine max, min, mean of lumped
  specific heat

Min Cp Avg Cp Max Cp
453.3 J/kgK 504.1 J/kgK 554.9 J/kgK
17
Transformer Cabinet Model
18
Cabinet Simulations

• Steady state conditions with a uniform air flow
  across the cabinets outer surface
• Transient simulation with no forced flow using
• Maximum specific heat
• Minimum specific heat

19
Cabinet Temperature Profile Steady State
20
Front View of Air Velocity ProfileSteady State
21
Side View of Air Velocity ProfileSteady State
22
Cabinet Simulation Results
23
Lumped Parameter Simulations (1/5)

• Objective
• Create a theoretical model of the EAB rooms
  thermal activity
• Provide an alternative solution method that will
  predict air heat up rate.
• Provide confidence in computational model.
• Allow an additional means of connecting the
  simulation results with the experimental results.

24
Lumped Parameter Simulations (2/5)

• In our current analytical approach, the room is
  reduced to two heat-storing masses, the cabinets
  and the air. From the basic equation for heat
  storage,
• two differential equations can be derived for the
  air temperature and cabinet surface temperature

and
25
Lumped Parameter Simulations (3/5)

• The two equations on the previous slide can be
  arranged in a heterogeneous linear system of
  equations, which can be solved simultaneously
  through matrix methods to yield

and
Where
is the second eigenvalue. (?1 0)
26
Lumped Parameter Simulations (4/5)

• To confirm simulation validity, geometric
  parameters were taken from SolidWorks model
• Pg heat generation 85700 W 292400 Btu/hr
• Mair mass of air in room 5657 lbs
• Mm mass of cabinets 1638000 lbs
• Cpa air heat capacity 0.241 Btu/lb F
• Cpm metal heat capacity .117 Btu/lb F

27
Lumped Parameter Simulations (5/5)

• Once all parameters are known, the constants C1
  and C2 can be determined from initial conditions
  (t 0). Initial conditions used
• Tair(0) 63.4 F
• and
• Tm(0) 181.3 F
• Once constants are known, equation for Tair
  104F can be solved for t, which may be used to
  determine Tm at that time

From SS simulation under normal operating
conditions
28
Analytical Solution Assumptions

• Uniform heat generation.
• The convection coefficient does not vary
  spatially.
• The convection coefficient is fairly constant
  over the temperature range.

29
Overall Approaches

• Three approaches
• Perform calculations by hand/in Excel
  spreadsheet.
• Model simplified version in FloWorks with
  cabinets lumped together.
• Use differential equation solver ODE45 in MATLAB

30
Analytical Solution Results

• Hand Calculations/Excel file (with h 6 W/m2 C
  122.4 Btu/hr ft2 F

31
Analytical Solution Results

• Simplified FloWorks Simulation (h calculated
  automatically by FloWorks a CFD package)

32
Experiment Overview

1. Goal
2. Approach
3. Experimental setup
4. Tests
5. Results

33
Experiment Goals

• Determine thermal conductivity (k)
• Benchmark the FloWorks CFD package using
  experimental results

34
Experimental Setup ( 1 /2 )
Fouriers Law
3.7in
2.5in
2.5in
35
Experiment Setup (2/2 )

• Aluminum steel blocks
• 2.5x2.5x3.7 in
• 200 W cartridge heater
• Approximately 95 Efficiency
• Block system
• Cartridge heater and thermocouples are covered
  with silicone grease to remove insulating effects
  of air

36
4. Tests

• Test 1
• Insulated aluminum block
• Power remains constant
• Determine the thermal heat generation and
  conductivity (k)

37
5. Results Test 1
Temp Deviation at 373.15 (deg C) k avg (W/mK) k_standard (W/mK) Error in k
0.28 190 200 5
38
Experiment Future Work

• Convection experiment using same setup
• Conduct testing with different materials
• Create FloWorks model with the same material and
  conditions to benchmark simulation results

39
Project Accomplishments

• Used computer simulation results to predict the
  heat-up rate of the EAB room.
• Normal Operating Conditions 63F
• Half flow single train failure 79F
• HVAC chiller failure 82.6F
• Total HVAC failure 19 minutes after total
  failure (104F)
• Derived equations to analytically calculate the
  heat-up rate using lumped parameter model.
• Heat-up rate 7 minutes after total failure
  (104F)
• Designed an experimental setup that can be easily
  compared with a Cosmos FloWorks CFD package.

40
Nuclear Engineering Initiative
QUESTIONS?

• Andron Creary, Kyle Bowzer, Brent Mayorga,
  Matthew King, Ryan Bigelow,
• George Campa, Jennifer Banegas, Matt Langston,
  Richard Colunga