You are on page 1of 104

Stratified Laboratory Thermal Energy Storage (LabTES) Tank Experiments: Sensible

Only and Sensible Augmented with PCM-Filled Tubes

A thesis submitted to the

Graduate School

of the University of Cincinnati

in partial fulfillment of the

requirements for the degree of

MASTER OF SCIENCE

In the Department of Mechanical and Materials Engineering

of the College of Engineering &Applied Sciences

by

KelseyLee H.Schafer

Bachelor of Mechanical Engineering, University of Cincinnati, Cincinnati, Ohio, 2013

Committee Chair: Michael J. Kazmierczak, Ph.D., Associate Professor


Abstract
The University of Cincinnati utilizes large scale stratified thermal energy storage (TES) tanks in

its water cooling system. The goal of this study is to investigate the potential benefits of an

affordable, easily manufactured and installed system incorporating phase change materials (PCM)

to augment the thermal capacity of these pre-existing stratified tanks. A 187-gallon laboratory TES

tank has been constructed and installed into the chilled water system at the University of

Cincinnati’s east power plant. A rack made of PVC with cylindrical copper tubes full of PCM

placed along its height was set in the center of the laboratory TES tank to test the effects of

augmenting the University’s TES tanks with PCM. The laboratory TES tank was tested first

without the PCM (sensible thermal storage) in order to gather reference data, and then was tested

with the cylinders of PCM (sensible and latent thermal storage) augmenting thermal capacity.

Separate charge and discharge half-cycle tests were run with flowrates between 0 and 2.5 GPM,

and then continuing half-cycle tests were run with a controlled flow rate of 2.5 GPM. Charge time,

discharge time, thermocline thickness, storage capacity, half figure of merit (FOM), and system

efficiency were calculated for each test and used to draw conclusions on tank performance. It was

found that the use of PCM in the laboratory stratified TES tank increased the thermal storage

capacity of the tank by approximately 10% while thermocline thickness remained nearly the same,

and system efficiency slightly increased.

i
ii
Acknowledgements

With special thanks to Joe Harrell, who funded this research, and to Dave Jeffries, Dan Pumphrey,

Jeff Finan, and Debbie Hausman, who helped purchase materials and assemble the LabTES tank

system and were great friends during the time I spend working at the Utility Plant. Also, special

thanks to my advisor, Dr. Michael Kazmierczak, who worked with me on both nights and

weekends to ensure the success of this thesis. Last I would like to thank my family, my friends,

and God for carrying me through these past 6 years of academia.

iii
Table of Contents
Acknowledgements ...................................................................................................................... iii
1 Introduction ........................................................................................................................... 1
1.1 Relevance of Study ................................................................................................................ 2
1.2 Objectives .............................................................................................................................. 2
2 Literature Review ................................................................................................................. 3
2.1 Stratified Tanks ..................................................................................................................... 3
2.2 Stratified Tanks in Use at University of Cincinnati ............................................................ 11
2.3 PCM/Heat Exchanger .......................................................................................................... 14
3 Thermal Storage Calculations ........................................................................................... 18
3.1 Thermal Storage for Un-Augmented Tank ......................................................................... 18
3.2 Thermal Storage for Tank Augmented with PureTemp 8 ................................................... 19
4 Experimental Design and Instrumentation ...................................................................... 21
4.1 Laboratory Stratified Thermal Energy Storage (LabTES) Tank ......................................... 21
4.2 Test Tubes and Rack ........................................................................................................... 23
4.3 Flow Circuit Accessories .................................................................................................... 25
4.4 Measurement and Instrumentation ...................................................................................... 27
5 Procedure & Data Reduction ............................................................................................. 31
5.1 Test Procedure ..................................................................................................................... 31
5.2 Data Analysis ...................................................................................................................... 34
6 Sensible (Un-augmented) Stratified TES Tank ................................................................ 38
6.1 Effects of Varying Flow Rate.............................................................................................. 39
6.2 Charge Time ........................................................................................................................ 40
6.3 Thermocline and Thermocline Thickness ........................................................................... 40
6.4 Integrated Capacity and System Efficiency ........................................................................ 45
6.5 Half FOM ............................................................................................................................ 47
6.6 Conclusions ......................................................................................................................... 49
7 Nominally Augmented Stratified TES Tank .................................................................... 51
7.1 Effects of Subcooling and Latent Heat ............................................................................... 54
7.2 Charge/Discharge Time....................................................................................................... 57

iv
7.3 Thermocline Thickness ....................................................................................................... 59
7.4 Thermal Capacity ................................................................................................................ 62
7.5 Conclusions ......................................................................................................................... 68
8 Overall Conclusions and Future Work ............................................................................. 71
8.1 Conclusions ......................................................................................................................... 71
8.2 Recommended Future Work ............................................................................................... 73
References .................................................................................................................................... 75
Appendices ................................................................................................................................... 78
Appendix A: Additional Figures ............................................................................................... 78
Appendix B: Matlab Codes ....................................................................................................... 83
Appendix C: Properties of PureTemp 8 .................................................................................... 94
Appendix D: Addition Thermal Capacity Calculations ............................................................ 96

v
1 Introduction

Stratified thermal energy storage (TES) tanks are tall, well-insulated tanks filled with a chosen

thermal energy storage material. These tanks are often installed in line with water cooling system

to augment the system’s capacity while reducing the operating costs. TES tanks are already used

commonly to augment cooling systems in commercial buildings to take advantage of off-peak

energy prices by storing energy at night during off-peak hours. This energy is then expended during

day time hours when energy prices are higher. The process of storing energy in a TES tank is

called charging, and the process of expending that energy is called discharging. Stratified tanks

which store hot and cold water in the same tank are extremely advantageous for thermal energy

storage due to their simplicity, excellent thermal efficiency, and overall cost effectiveness.

For sensible storage, water is widely considered the best storage material to use in these tanks.

However, in recent years, there has been rising interest in latent heat storage methods, which take

advantage of the large amount of energy expended or stored or released during the phase change

of a storage material. For latent heat storage to be used, the storage material must have a freezing

point that is within the operating temperature range of the system in which the TES tank is to be

installed. For example, a typical chilled water cooling system operates between 40℉ and 50℉.

Therefore, water, which freezes at around 32℉, could not offer any latent heat storage in such a

system. Recent advances in materials engineering have produced phase change materials (PCM)

specifically designed to have the desired thermal properties required to store latent heat energy at

desired temperatures. These PCMs come in a variety of forms such as paraffin waxes, plant-based

oils, or even water-glycol mixtures. These technological advances make PCMs a potential tool for

augmenting thermal storage. One way to integrate PCM with a stratified TES tank is to place

1
sealed containers filled with PCM into the tank, which would theoretically behave much like ice

cubes in a drink. This thesis will examine the feasibility and possible benefits of such an approach.

1.1 Relevance of Study

At the University of Cincinnati there are two large stratified chilled water TES tanks, which store

chilled water at night for use during the day in the warmer months of the year. It is desirable to

increase the thermal energy storage capacity of these tanks to meet the future cooling needs of the

university. It is vital to know the effects that integrating PCM into these TES tanks would have on

over-all system performance. Currently, we’ve found no existing literature studying these effects

in cold storage tanks. Because integrating PCM could have unforeseen effects on the TES tanks at

University of Cincinnati as well as the University’s water cooling system, it is reasonable to

conduct preliminary testing with PCM on a smaller laboratory TES tank to reduce risk.

1.2 Objectives

The objectives of the Mini-Thermal Energy Storage Tank are as follows:

1. Design and build a laboratory stratified TES (LabTES) Tank with necessary instrumentation.

Demonstrate and document thermocline development in LabTES tank for different flow rates.

(Obtain reference data without PCM tubes.)

2. Use PCM to augment LabTES tank capacity using a simple PCM-filled tube arrangement.

3. Measure the performance enhancement for tank when using PCM and determine any negative

effects on thermocline or tank stratification, if any.

4. Determine the feasibility of using a similar method to augment stratified TES tanks at

University of Cincinnati.

2
2 Literature Review

In order to understand how to best augment University of Cincinnati’s thermal energy storage

tanks, a study was made of the existing literature on the following subjects: (1) stratified TES

tanks, (2) the thermal energy storage tanks already built on campus at the University of Cincinnati,

and (3) the phase change materials to be incorporated in the tanks. An overview of the past research

in stratified tanks gives a basic understanding of the system parameters and requirements to build

the mini-TES tank, and an analysis of the TES tanks currently in use at U.C. yields the more

detailed specifications for building our experimental set-up. Last, a study of research in phase

change materials gives an understanding of what kind of PCM should be applied to augment the

tanks. This literature review summarizes the current body of knowledge on these topics.

2.1 Stratified Tanks

In simplest terms, a naturally stratified water tank is a vessel in which the more buoyant warmer

water has floated to the top while the cooler water has remained at the bottom, separated by a

thermocline. When studying stratified tanks, it’s important to take into account the fluid mechanics

effecting stratification as well as how the thermocline behaves. This will ultimately affect the

efficiency of the tank.

Stratified Tank Research

In a study of chilled-water storage [1], Wildin and Truman conducted several experiments with

large and scale model storage tanks. Some of the tanks used a membrane or walls or baffles to

achieve stratification, while others use natural stratification by buoyant forces. It was concluded

that naturally stratified tanks are not only more simple to design and operate but also more

thermally efficient, about 90% efficient.

3
Wildin followed this study with a report [2] on diffuser design, which described simple

mathematical formulas for designing the length of the diffuser and the inlet opening height based

on the Reynolds number and Froude number, respectively. The procedure relates the flow rate per

unit length of the diffuser to the temperature difference (∆T) and Reynolds number of the fluid and

then uses the flow rate per unit length and the desired Froude number to calculate the height of the

slots in the diffuser where water enters the tank. Results of this report also compared the figure of

merit (FOM) as a function of tank outlet temperature for a small radial diffuser, a large radial

diffuser, and an octagonal diffuser and found that the octagonal diffuser provide the highest FOM.

These findings supported the octagonal diffuser as the better option for increasing tank efficiency.

Wildin’s studies became the basis of current ASHRAE standards for stratified tanks.

In 1989, Truman and Wildin [3] collaborated to produce a finite difference model for stratified

tanks which accounted for the conduction and convection between the wall and floor of the tank

and the liquid as well as mixing and the heat loss from the tank to the surroundings. Two-

dimensional heat conduction was used for modeling the tank walls and floor with a one-

dimensional vertical temperature distribution was used for the liquid in the tank. This model called

STRATUNM accurately predicted these temperature distributions as well as the available

discharge capacity of the tank.

Zurigat et al [4] took this modeling study a step further, comparing six different models by Sharp,

Han and Wu, Cole and Bellinger, Cabelli, Wildin and Truman, and himself. In a series of tank

charging experiments, Zurigat examined how well each of these models agreed with experimental

data. Each model was one-dimensional and assumed a constant temperature for water entering the

tank. Results of this study concluded that the models of Cole and Bellinger, Zurigat, and Wildin

and Truman are superior in accuracy to others though they are not as computationally efficient.

4
Zurigat also concluded that his own model was best in accuracy because it incorporates inlet

mixing correlations in order to make more accurate predictions when mixing is significant. Other

models struggled to predict temperature distributions in areas of the tank where mixing was severe.

Wildin’s next study in 1991 [5] of flow near the diffuser inlet focused on the effect of the Reynolds

number (Re) and Froude number (Fr) on the gravity currents traveling through a stratified tank and

on the thermocline. The Froude number was maintained below 2, and eight different charging tests

were run each with a Reynolds number in the range of 159 to 633. Results showed that for the

tested values of Fr, values of Re above 200 produced a thicker thermocline. Also, Re values above

400 produced significant mixing in the tank, increasing the slope of the temperature gradient and

rendering stratification ineffective. Gravity currents were found to traverse the tank with greater

velocity as Re increased and were found inside the thermocline as well as below it. The study also

found that the difference between the temperature of the chilled water entering the tank and the

temperature of the water at the inlet increased with increasing Reynolds number, decreasing the

effective storage.

Much of research up to this point in time was done using scale model stratified tanks. Bahnfleth

[6] later recognized the lack of data for full scale existing stratified tanks and did further research.

In 1998, Bahnfleth published a study on the thermal performance of a 1.47-million-gallon

cylindrical tank with radial parallel plate diffusers and a constant inlet flow rate. A series of charge

and discharge cycles were analyzed, and values for integrated capacity, figure of merit, and lost

capacity were calculated. Results showed that the uncertainty in both the integrated capacity and

figure of merit were highly influenced by the accuracy of the flow rate measurement. It was

determined that even a fairly accurate flow meter would not be sufficient to improve those

uncertainties, making measures of integrated capacity and figure of merit less meaningful. Instead,

5
it was found that especially for taller tanks, the measurement for lost capacity involved less

uncertainty.

In 2001, Stewart [7] challenged the current ASHRAE guidelines, which were based mostly upon

small cylindrical tanks of 35,000 gallons or less, and investigated whether these guidelines were

too stringent when applied to larger tanks. Stewart collected data from five different stratified tanks

already being used in industry. It was found that four out of the five tanks were never completely

discharged but rather discharged only enough to meet the cooling load and then charged again (i.e.

partially discharged). For this reason, the figure of merit could not be calculated. However, the

thermocline thickness could be measured, and this could be used to approximate the efficiency of

the tanks. Results showed that for larger tanks, a Reynolds number of up to 6000 would not

significantly increase thermocline thickness. The study also suggested that further research into

charging and discharging with a non-constant flow rate could be meaningful.

Around the same time, Musser and Bahnfleth [8, 9, 10] followed up his study of large stratified

tanks with a parametric study of inlet diffuser performance. A CFD model was created and

implemented on two large scale tanks. Modeling focused on the lower region of the tanks near the

inlet where the thermocline develops. The CFD model assumed a uniform inlet velocity profile

and laminar flow. Both the effects of turbulence and inlet temperature variation were considered

in the creation of the model. Comparison of experimental thermocline data from these tanks with

thermocline predictions based on the CFD model validated the laminar CFD model.

This CFD model was used in a series of numerical experiments Musser and Bahnfleth [8, 9, 10]

performed in order to quantify the effects of six diffuser and tank design parameters on the inlet

thermal performance. Bahnfleth’s tests included 2k factorial experiments, dimensional analysis,

6
and analysis of sixteen tests on large scale stratified tanks. This study led to the development of

first-order regression models for thermocline thickness and equivalent lost tank height (similar to

lost capacity) based on the inlet Richardson number, the ratio of diffuser diameter to diffuser

height, and the ratio of diffuser diameter to tank diameter. Results of this study found the ratio of

diffuser radius to tank radius and the inlet Richardson number to be the most significant flow

parameters for tanks with slot-type diffusers and found Reynolds number to be of secondary

importance. The study also concluded that larger Richardson number and smaller inlet slot height

resulted in improved performance.

Thermal Efficiency and Figure of Merit (FOM)

Thermal efficiency is the ratio of the energy that is able to be taken out of a storage tank to energy

put in, basically the ratio of the discharging capacity to the charging capacity. The primary measure

of thermal efficiency in a stratified tank is the discharge capacity, meaning the amount of thermal

energy the tank is able to store and provide to the system it augments. This is exemplified by

Bahnfleth’s measure of system efficiency. Another common measure of efficiency is the figure of

merit (FOM), which is the ratio of integrated discharge capacity for a given volume to the ideal

capacity that could have been withdrawn in the absence of mixing and losses to the environment

[6]. The figure of merit is usually calculated for the entire process of charging and discharging.

However, in his experiments Bahnfleth also used the half-cycle FOM, which is calculated for only

one discharge or one charge process. Lost capacity is another more practical measure of tank

performance. It is defined by Bahnfleth as the capacity that cannot be removed from the tank due

to an outlet temperature limitation [6].

According to the study by Zurigat, Ghajar, and Maloney [4] on one-dimensional modeling of

stratified tanks, there are four major causes of loss of stored energy and thermal efficiency. These

7
are heat gain from the water in tank from the surroundings, thermal diffusion through the

thermocline, convection currents in the water causing mixing, and mixing introduced at the inlet

during charging and discharging.

Several aspects of heat gain to the water in the tank must be considered. The top surface of the

water is left exposed to atmospheric pressure and will suffer convective gain as will the outer walls

of the tank. There is also gain from the water through the tank walls themselves. Heat gain through

the floor of the tank is of lesser concern than these but should still be considered. These problems

are fairly easily minimized by properly insulating the walls and floor of the tank as well as any

pipes leading to and away from the tank. Building tanks underground offers very effective

insulation though the precise amount of insulation depends heavily on the soil moisture content.

Another way to minimize these heat loss effects is to minimize the ratio of the surface area of the

tank to its volume. The EPRI Stratified Chilled-Water Storage Design Guide [11] suggests that a

cylindrical shaped tank offers the optimum aspect ratio of surface area to volume.

When studying thermal storage, it is important to consider the effect that the thermocline has on

the efficiency of the storage system. The thermocline separates cooler, denser water below from

the warmer, less dense water above it. This separation is called stratification. The thermocline also

acts as a barrier to mass and heat transfer, meaning, that the mixing of water above the thermocline

has less effect on the water below the thermocline. The thermocline is created by a gravity current

which is initiated by the temperature difference between the water entering the tank and the water

at the inlet. The Froude number and Reynolds number determine the thickness of this gravity

current which then determines the thickness of the thermocline. As time passes, diffusion between

the warm and cool water increases the thickness of the thermocline. Heat gain from the tank walls

mentioned above also contributes minorly to the thickening of the thermocline. Increasing

8
thermocline thickness decreases the amount of usable chilled water in the tank, thus reducing the

discharge capacity and efficiency.

Mixing at the inlet and in the rest of the tank also causes increased thermocline thickness. Mixing

is mass and heat transfer that occurs between the warmer and cool water, effectively eliminating

stratification. It caused by undesirably high Reynolds and Froude numbers as well as by unwanted

convection currents and vertical fluid momentum. The dimensionless Froude number (Fr) is the

ratio of the inertial forces to the gravitational forces on a fluid element. Wildin [5] suggested a

value of Fr less than 1 because Fr is less than 1 a gravity current forms at the inlet and helps to

prevent mixing (flow near inlet). Reynolds number is the ratio of inertial to viscous forces on a

fluid element. A low Reynolds number and Froude number and reduced mixing can be achieved

through proper design of diffusers.

Storage Tank and Diffuser Design

Tank dimensions and geometry and diffuser design must both be carefully considered in the

design of a thermal energy storage system. Thermal energy storage tanks are most commonly

square or cylindrical in geometry, and single-tank systems are preferred due to simplicity and

decreased surface area to volume ratio. Diffusers are pipe systems at the bottom and top of the

tank that distribute the inlet and outlet flow. This section will expand upon the details of tank sizing

and diffuser design.

The dimensions of a storage tank are especially important for reducing mixing and preserving

usable storage capacity. As discussed before, a vertical cylinder is the most effective geometry to

minimize the surface area to volume ration. The walls of the tank should be vertical because any

9
curvature in the tank walls will cause unwanted horizontal motion in the flow through the tank.

That means different directional flows will mix in the tank.

Tank depth also plays an important role. According to the EPRI Chilled Water Storage Guide

asserts that tanks less than 5 feet tall will significantly decrease usable tank volume. The tank used

in this experiment is approximately 5.3 feet tall. The volume of the tank should be chosen based

upon the cooling needs of the system. However, tank height becomes less influential on usable

volume as tank height increases higher than 6 feet. And significantly increasing the surface area at

the top of the tank exposes more water to ambient air, which increases heat gain. Therefore, as

tank size and volume increase, tank insulation must also be improved.

Diffusers are commonly made up of a system of pipes arranged in a labyrinthine, hexagonal, or

radial geometry. These pipes have rectangular slot holes at intervals along the pipes to allow exit

and entrance of inlet and outlet water at distributed places in the tank. Labyrinthine diffusers are

made up of a complex, matrix-like system of pipes spanning the bottom and top areas of the tank.

Radial diffusers are composed of smaller diameter pipes extending outward from a main inlet pipe

in the center of the tank. Hexagonal diffusers are made up of pipes arranged in the configuration

of one or two concentric hexagons. The diffusers used in this experiment are hexagonal.

A properly designed diffuser at the inlet and outlet of a thermal energy storage tank distributes the

flow of incoming and outgoing flow to reduce mixing and unwanted currents in the tank and helps

to ensure natural stratification within the tank. According to ASHRAE standards diffusers should

be designed to ensure a Froude number of less than 2. And a desirable Reynolds number is under

1,000 [2]. The design of the diffusers in this experiment was based on this Reynolds and Froude

10
number criteria. The diameter of the hexagon geometry and height of the rectangular slotted holes

along the diffuser pipes were calculated using a method prescribed by Wildin [2].

2.2 Stratified Tanks in Use at University of Cincinnati

One of the goals of this study is to augment the capacity of the large TES tanks at University of

Cincinnati power plants. In order to do this, one must first be familiar with the structural

specifications and normal operation of the University TES tanks. This section offers a brief

description of the east and west campus thermal energy storage tanks.

There are two large thermal energy storage tanks at the University of Cincinnati, one located

underground at U.C. east campus power plant and another located under the training football field

on the university campus. Both tanks run within the larger water system that runs throughout the

entire university. Just like the laboratory TES tank built for this experiment, these two large tanks

use the operating temperature range supplied by the chillers at the east campus power plant. Unlike

the laboratory TES tank, however, the larger tanks are not usually charged completely, or

discharged completely. Instead, the water in the tanks if chilled just enough to accommodate the

load required by the university’s larger water system on any given day. This means that in the

winter or colder months, when the cooling load at the university is much smaller, the two tanks

are operated with very small temperature differentials or not at all. The tanks are primarily used

during the summer months.

The tank at the east campus power plant was built in 1997. Reference Figure 2.1 and 2.2 for a side

and top view schematic of the east campus power plant tank. This tank is 90 ft. long by 82 ft. wide

11
Figure 2.1 Side View East Campus Power Plant Tank

Figure 2.2 Top view East Campus Power Plant Tank

12
Figure 2.3 Side View Campus Football Field Tank

Figure 2.4 Top View Campus Football Field Tank

13
by 24 ft. deep and is filled with approximately three million gallons of water. This gives the tank

an approximate thermal storage capacity of 20,932 Ton-hours. The tank under the campus football

field was built in 2012. Reference Figure 2.3 and 2.4 for a side and top view schematic of this tank.

This tank is 167 ft. long by 126 ft. wide by 27 ft. deep and is filled with approximately four million

gallons of water. This gives the tank an approximate thermal storage capacity of 27,909 Ton-hours.

2.3 PCM/Heat Exchanger


Also critical to the performance of the tank are the thermal properties of the phase change material

and its chemical compatibility with the materials making up the heat exchanger. Phase change

materials (PCM) are defined as materials that take advantage of the latent heat involved in

solidifying liquefying, evaporating, or condensing in order to store or release large amounts of heat

energy. The PCM used in this experiment is called PureTemp 8 and is supplied by Entropy

Solutions, Inc. It is a non-toxic, renewable, and biodegradable PCM made from vegetable

products. PureTemp 8 was chosen for this experiment based on its theoretical fusion temperature,

its high specific and latent heat values, its chemical compatibility with other materials in the

experimental set up, and its exceptional lifespan of over 20,000 thermal cycles without thermal

degradation. This section will give a short history of the available literature on phase change

materials and a more detailed description of the properties of PureTemp 8.

Phase Change Material Research

Over the past forty years, rising climate change and energy consumption has prompted scientific

advances in energy storage technologies which improve energy and cost efficiency. Shortages of

petroleum products in the 1970s brought on the investigation of storage materials other than ice.

Feasibility studies and conferences continued into the late 1990s and early 2000s when thermal

storage applications for PCMs were more heavily researched.

14
In 2008, Dr. Harald Mehling and Prof. Luisa Cabeza collaborated to write Heat and Cold Storage

with PCM [12], a comprehensive book detailing the basic concepts of thermal storage and design

of latent heat storage systems, experimental methods of determining the thermal properties of

PCMs, and the many possible applications of phase change technology. This book draws on

Mehling and Cabeza’s past work in the area of thermal storage as well as on past work in thermal

storage by ASHRAE, Wildin, and others previously mentioned in this thesis.

Thermal and Chemical Properties of PCM

PCMs are thermal storage materials that are carefully developed to have specific thermal properties

which allow them to take advantage of latent heat energy at a temperature that is within the

operating temperature range of the system they are used in. Simply put, PCMs are designed to

have a certain desired fusion temperature. This desired fusion temperature should be greater than

the chilled water temperature in the cooling system but less than the warm return water

temperature. In order to change phase and store latent energy efficiently in a system, there are five

important thermal properties of the PCM that must be considered: (1) fusion temperature, (2)

thermal conductivity, (3) specific heat, (4) latent heat, and (5) chemical compatibility. These are

explained further in this section.

The temperature of fusion (𝑇𝑓 ) of a material is the temperature at which the material changes phase.

This value is often determined using a T-history test or conventional calorimetry methods like

differential thermal analysis or differential scanning calorimetry [13]. In the case of such as

experiment, PureTemp 8 was selected because it is designed to change from a liquid to a solid at

a fusion temperature of 8℃ = 46.4℉. However, this is an ideal value. In reality, the fusion

temperature of a PCM may vary slightly from sample to sample of the same PCM. This is because

most PCMs are not pure materials but rather mixes of materials. Because the fusion temperature

15
may vary from sample to sample, it is important that the system have a larger operating temperature

range.

Thermal Conductivity (k), according to Mehling and Cabeza [12], is the ability of a material to

transport heat while the material itself does not move. The thermal conductivity of a material

determines the rate at which heat transfers to and from a material. A higher rate of heat transfer

will ensure that both latent and sensible energy will be transferred to and from the PCM more

rapidly. Therefore, a PCM must have a relatively high thermal conductivity to accompany a high

latent heat value.

Specific Heat (c) is the amount of thermal energy that can be stored in a unit of mass for every

degree of temperature change during a change in temperature that is not at the fusion temperature.
𝐽
Specific heat is expressed in 𝑔℃ . Latent heat (L) is the amount of thermal energy that can be stored

or released in a unit of mass during the process of phase change. Consequently, latent heat is

expressed in units of J/g. The specific heat of a PCM is usually significantly less than that of water,

which is commonly used as a thermal storage medium. However, the high value of latent heat in a

PCM is meant to make up for this deficiency. Thus it is extremely important for a PCM to have a

high latent heat value in order to increase the overall storage capacity of the system. The specific
𝐽 𝐽
heat of PureTemp 8 are 2.15 for liquid PCM and 1.85 for solid PCM, and value of latent
𝑔℃ 𝑔℃

𝑗
heat is 178 𝑔. For more information about the properties of PureTemp 8, see Appendix C.

The term chemical compatibility refers to the ability of a PCM to interact with other system

components without causing material degradation or other undesirable effects. PureTemp 8

specifically is less chemically compatible with components made of polypropylene or

16
polyethylene plastics as well as some types of rubber sealants but is non-reactant with metals.

Therefore, it’s vital to consider what materials the PCM comes in contact with. For this reason,

the heat exchanger tubes used to contain the PCM were made of either metal or PVC.

17
3 Thermal Storage Calculations

This chapter offers theoretical calculations for ideal thermal storage capacity of the un-augmented

and augmented laboratory TES tank. Simple thermodynamic analysis was used to calculate the

amount of thermal energy stored in the laboratory TES tank assuming no flow of water is running

through the tank.

The ideal charging situation for the thermal storage tank implies four assumptions: 1. The entire

tank cools from the initial temperature to the final temperature in the time it takes for the

thermocline to pass once from the bottom of the tank to the top (one charge cycle). 2. The initial

temperatures of both the water and the PCM in the tank are equal to return water temperature.

𝑇𝑊𝑖 = 12℃ and 𝑇𝑃𝑖 = 12℃. 3. As chilled water enters the tank, both the water and the PCM cool

at roughly the same rate to the temperature of the chilled supply water, 𝑇𝑊𝑓 = 6℃ and 𝑇𝑃𝑓 = 6℃.

Thus, the total change in temperature ∆𝑇 = 6℃. 4. There are no energy losses from the water and

PCM in the tank to the environment or the tank itself. 5. The inlet water temperature is constant.

These assumptions also apply for the discharging process with the exception that the initial

temperature is equal to the chilled water temperature, and the final temperature is equal to the

return water temperature. 𝑇𝑊𝑖 = 6℃, 𝑇𝑃𝑖 = 6℃, 𝑇𝑊𝑓 = 12℃, and 𝑇𝑃𝑓 = 12℃ for the discharge

cycle. Also, for the discharge cycle, the thermocline travels from the top of the tank to the bottom.

3.1 Thermal Storage for Un-Augmented Tank


Because the melting temperature of water is much lower than the operating temperatures of this

experiment, there is no latent component of energy storage in the un-augmented tank. Therefore,

the total thermal storage of the un-augmented tank is equal to the sensible heat storage of the water

in the tank.

18
𝑄𝑇𝑎𝑛𝑘 = 𝑄𝑊 = (𝜌𝑉𝑊 𝐶∆𝑇)𝑊 (Eq. 3.1)

𝑘𝑔 𝑘𝐽 1 𝑘𝐽
𝑄𝑊 = 1000 3
∙ 0.7072 𝑚3 ∙ 4.204 ∙ 6℃ ∙ (Eq. 3.2)
𝑚 𝑘𝑔℃ 12660 ton ∙ hr

𝑄𝑊 = 1.409 ton ∙ hr (Eq. 3.3)

3.2 Thermal Storage for Tank Augmented with PureTemp 8


The total ideal storage capacity for the tank augmented by the PCM is calculated as the sum of the

ideal capacity of the water plus the ideal capacity of the PCM. The ideal capacity of the PCM is

calculated below as the sum of the sensible storage of liquid PCM, the latent heat, and the sensible

storage of the solid PCM, where the theoretical phase change temperature of PureTemp 8, 𝑇𝐹 =

8℃. First, the capacity of the PCM is calculated as follows:

𝑄𝑃 = 𝜌𝑃 𝑉𝑃 [𝐶𝑃𝑙 (𝑇𝑃𝑖 − 𝑇𝐹 ) + 𝐿 + 𝐶𝑃𝑠 (𝑇𝐹 − 𝑇𝑃𝑓 )] (Eq. 3.4)

𝑘𝑔 𝑘𝐽 𝑘𝐽 𝑘𝐽 1 𝑘𝐽 (Eq. 3.5)
𝑄𝑃 = 860 ∙ .01094 𝑚3 [2.15 ∙ 4℃ + 180 + 1.85 ∙ 2℃]
𝑚3 𝑘𝑔℃ 𝑘𝑔 𝑘𝑔℃ 12660 ton ∙ hr

𝑄𝑃 = 0.1428 ton ∙ hr (Eq. 3.6)

The sensible storage in the volume of water filling the rest of the tank is calculated in the same

way as before. Note that the value for volume used here accounts for both the presence of the PCM

and the heat exchanger.

𝑘𝑔 𝑘𝐽 1 𝑘𝐽
𝑄𝑊 = 1000 ∙ 0.705 𝑚3 ∙ 4.204 ∙ 6℃ ∙ 12660 ton∙hr = 1.405 ton ∙ hr (Eq. 3.7)
𝑚3 𝑘𝑔℃

19
Thus, the ideal storage capacity of the augmented tank is as follows.

𝑄𝑇𝑎𝑛𝑘 = 𝑄𝑊 + 𝑄𝑃 = 1.547 ton ∙ hr (Eq. 3.8)

With a volume fraction of 1.527%, the ideal improvement in thermal storage is calculated to be

9.821%. The storage capacity of the copper heat exchangers themselves, the PVC rack supporting

the heat exchangers, and the PVC distribution rings in the tank is assumed to be negligible as it is

less than .5% of the total thermal capacity of the tank. Refer to Appendix D for this calculation.

20
4 Experimental Design and Instrumentation
The experimental set-up for the LabTES experiments was designed and built according to

standards set forth in the EPRI Stratified Chilled-Water Storage Design Guide [11], ASHRAE

standards, and standards described in the Design Guide for Cool Thermal Storage from the

American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. [14]. The

system consists of a stratified thermal storage tank with flow distribution rings, a rack of tubes

filled with PCM, the necessary piping, valves, pump, and instrumentation, and a data acquisition

program to automatically operate the system.

4.1 Laboratory Stratified Thermal Energy Storage (LabTES) Tank

The main test section of the LabTES tank set-up (Figure 4.1) is a 66-inch-tall plastic storage tank

29 3⁄16 inches in diameter, which holds 187 gallons of water. The tank was designed as a small-

scale simulation (with the same rise rate range) of the larger commercial TES tanks used in the

University of Cincinnati Central Power Plant. The tank is open at the top, exposed to atmospheric

pressure. It has an inlet and outlet at the top of the tank (upper inlet and outlet) and at the bottom

(lower inlet and outlet).

Two circular, PVC distribution rings, each with twelve rectangular slots, are mounted inside the

LabTES tank connected to the inlets/outlets both upper and lower. The process for designing the

height of these slots and the diameter of the rings is given in by Wildin [5]. This design process is

based on the criteria that the dimensionless Froude number is less than 1. The slots are designed

to be .5 inches wide and .75 inches in height and face alternatingly inward and outward from the

center of the tank. The diameter of each distribution ring is one half the diameter of the tank. These

21
Figure 4.1 Photograph of LabTES System

Chilled Water
Outlet
Return Water
Inlet (Return 1)

Chilled Water
Inlet (Supply) Return Water
Outlet

Figure 4.2 Distribution Rings Drawings (left) and Photograph (right)

Slots: .5” wide X .75” tall

22
distribution rings prevent turbulent flow in and out of the TES tank to prevent unwanted mixing

of the chilled water at the bottom of the thermocline and the warmer water at the top. The rings

also ensure that the flow rate into the TES tank and the flow rate out are equal, preserving a

constant volume of water in the tank. Refer to Fig. 4.2 for a drawing and photograph of the

distribution ring (left) and an overhead view looking through the open top down into the tank

(right).

When the system charges, chilled water at about 6 ℃ flows from the utility plant chilled water

supply line into the LabTES tank through the lower inlet while warmer water at the top of the

thermocline flows out through the upper outlet and is pumped back into the utility plant return

water line. During charging, the lower outlet and upper inlet are closed. When the system

discharges, warmer water at about 12 ℃ flows from a separate utility plant return water line into

the LabTES tank through the upper inlet, and chilled water at the bottom of the thermocline flows

out through the lower outlet and is pumped back into the utility plant return water line. The upper

outlet and lower inlet are closed. Figure 4.3 shows the flow path of water through the tank.

4.2 Test Tubes and Rack

A ladder-like, vertical rack was built from PVC to hold up to eighteen tubes of PCM. For the

standard test case, the nine tubes are mounted horizontally one above the other on the rack about

6 inches apart in height. An RTD is inserted into the center of each tube to measure the temperature

of the PCM. At first, for the purposes of this experiment, different materials were considered for

the tube wall. One set of nine tubes was made of copper, another of PVC, and another of

Aluminum. Ultimately, the copper set was chosen over the aluminum an PVC for its high thermal

23
Figure 4.3 Flow Paths of Water Through LabTES Tank in Charging and Discharging
Charging Flow Path Discharging Flow Path

Closed Inlet/Outlet Valve Open Inlet/Outlet Valve Water Flow Path

Figure 4.4 Test Rack


(left) Two-inch diameter PCM-filled
tubes in PVC, copper, and Aluminum
with RTDs. (right) Rack of nine PCM
tubes to be inserted into mini TES tank.
RTD

24
conductivity. Each coper tube is filled with a fatty acid type phase change material called

PureTemp 8 provided by Entropy Solutions Inc. Figure 4.4 shows the tubes and rack.

4.3 Flow Circuit Accessories

Piping, Valves, and Fittings

The TES tank has two inlets: The chilled water supply and the warm water return. The pipes to

these inlets are ½” copper. At both return and supply inlets are a ½” full port ball shut off valve, a

½” Y type strainer, a ½” pressure regulator set to allow 60 psig, a pressure gauge (0-100 psig), a

½” CWR control valve, a manually operated shut off valve, and an RDT to measure the inlet

temperature (Figure 4.5 Items 1-6). At both outlets of the tank, there in an outlet temperature RTD

and a 2” actuated PVC valve (Figure 4.5 Item 10). All pipes exiting the tank are 2” PVC.

Collection Tank

Water discharged from the TES tank flows into the 470-gallon rectangular collection tank (Figure

4.5 Item 11), where the flow rate is measured. Flow into the collection tank enters at the bottom to

prevent any disturbance at the surface of the water where the magnetostrictive level transmitter

(Figure 4.5 Item 12) floats. The tank sits on a square, wooden base 14.125 inches off the floor.

This tank is either manually or automatically drained after each charge test and each discharge test

to prevent overflow.

Pump

To empty water from the collection tank back into the power plant’s water supply, a Pump-Goulds

eSV Model 3SV12GF4C60 Multi-Stage Pump (Figure 4.5 Item 14) is located just after the

isolation valve (Figure 4.5 Item 13). The pump is rated at 3 HP and 3500 RPM. It runs with a flow

rate between 3 and 10 GPM. The pump does not run continuously. Its purpose is to drain the

25
Figure 4.5 LabTES System Diagram
6
3 5 12
10
1 12 4

Return 1 7

Return 2
11
8

16
13
15

Supply 9 14

1. ½” full port ball shut off valve 6. Inlet and outlet RDTs 12. Magnetostrictive level
2. ½” Y type strainer 7. Upper distribution header transmitter
3. ½” pressure regulator allows 8. Test section (thermal 13. Pump isolation valve
60 psig storage tank) 14. Pump
4. Pressure gauge (0-100 psig) 9. Lower distribution header 15. Pressure gauge (0-100 psig)
5. ½” CWR control valve & 10. 2” actuated PVC valve 16. Pump check valve &
manual shut off valve 11. 470-gallon collection tank isolation valve

Flow path of water during charging process from upper inlet (Return 1) to lower outlet
Flow path of water during discharging process from lower inlet (Supply) to upper outlet

26
collection tank when the water level rises to the set point height programmed in the programmable

logic controller. After the pump is a pressure gauge (0-100 psig), a check valve, and another

isolation valve (Figure 4.5 Items 15-16). Because the pipe exiting the pump flows vertically, the

check valve is necessary to prevent back flow due to gravity. Upon leaving the pump, water is

returned to the chilled water return line.

4.4 Measurement and Instrumentation

Temperature

Temperature data was taken using 120-inch-long hermetically sealed RTDs (Figure 4.5 Item 6)

from Omega. RTDs were calibrated against a constant temperature water bath. RTDs were placed

at the inlets and outlets of the LabTES tank as well as other locations inside the tank. Inside the

LabTES tank, an RTD was placed in the water at the same level of each PCM tube to display the

thermocline in the tank, and an RTD was placed inside each of the nine PCM tubes. These RTDs

inside the tank provide data used to measure the thermocline and to calculate other important

performance measures for the tank.

Pressure

An analog pressure gauge (0-100 psig) was placed just upstream of the CWR control valve on both

the chilled water supply line and the return water supply line. Pressure regulators placed just

upstream of the gauges ensure the pressure in the system remains at around 75 psig. By keeping

the system pressure constant, we eliminated pressure fluctuations in the supply line as a factor in

the experiment. There is also a pressure gauge located just downstream of the pump. For further

assurance, the CWR control valve (Figure 4.5 Item 5) automatically adjusts the inlet flow rate in

response to small changes in pressure to produce a constant flow rate.

27
Volumetric Flow Rate

The volumetric flow rate through the system is designed to be 2.5 GPM or less and is measured

by a Jogler Magnetostrictive Level Transmitter, Model LGT-6000 mounted inside the collection

tank (Figure 4.5 Item 12). The magnetostrictive level transmitter consists of a transmitter head,

sensor pickup, waveguide probe, and the waveguide float. The metal float floats at the surface of

the water in the tank and travels up and down along the vertical length of the waveguide probe.

The transmitter head sends a current through the waveguide probe creating a magnetic field that

interferes with the magnetic field of the float. From the reaction of the float’s magnetic field to

that of the probe, a torsional force is created. The length of time between the sending of the current

from the transmitter to the time the transmitter detects the torsional force is proportional to the

voltage output of the transmitter. The transmitter reads the water level every .1 seconds, and the

flow rate in the system is calculated as the difference in water level divided by the difference in

time. The accuracy of the level transmitter is +/- 0.015 inches. A 4-20 mA output signal sends

water level data to the programmable logic controller (PLC).

To supplement the above method of measuring the volumetric flow rate, a measuring stick was

placed in the main test section to measure the water level in the test section. During testing, this

water level is monitored and recorded. The water level in the tank remains constant. Therefore, the

flow into the tank and the flow out are equal.

Programmable Logic Controller

The programmable logic controller (PLC) is a graphical user interface created using RS Logics

5000. It is used to set the charging and discharging flow rates, control the pump and valves, and

collect and record all data for the experiments. The program features a time clock with hours,

minutes, and seconds to which all the data is synced. It also offers separate manual and automatic

28
Figure 4.6 PLC System Controller and Data Logger Screens

Home Screen Pump Control Screen

Data History Screen Tank Temperature Screen

29
modes for the pump. A discharge process and a charge process are programed into the PLC. When

charging the tank, the PLC sends commands to open the lower inlet valve and the upper outlet

valve, setting a user-defined set point flow rate. In contrast, when discharging the tank, the PLC

sends commands to open the upper inlet valve and the lower outlet valve, again setting the flow

rate of the inlet valve to the user-defined set point. The graphical user interface of the PLC features

a home screen where the user input is entered, a maintenance screen for managing any maintenance

to the system, a data table screen, and a temperature data screen, which shows an image of the TES

tank complete with temperature indicators (See Figure 4.6). The PLC stores the data in an excel

sheet at the end of each day and sends it to the Utility Plant network.

30
5 Procedure & Data Reduction

Figure 5.1 shows the basic procedure for each half cycle charge and discharge test ran in this study.

First, these charge and discharge tests were run without any PCM or copper tubes mounted in the

tank (i.e. a sensible configuration). Then these charge and discharge tests were run with nine

copper tubes filled with PCM mounted on the rack in the center of the tank (i.e. a nominally

augmented configuration). The steps taken to perform each test and the calculation performed for

data analysis of each test are explained below.

5.1 Test Procedure


Step 1: Prepare System Configuration

Looking at the PLC control screen, the operator checked the campus chiller’s output temperature

and the system’s return water temperature conditions. The difference between these two

temperatures was the operating ∆T. The operator ensured that the operating ∆T was greater than

8℉ before beginning any test. Next, the operator mounted the number of tubes (configuration)

desired for test in the appropriate sections of tank. For sensible tests, no tubes of PCM were used.

For nominally augmented tests, nine 2” diameter copper tubes (one tube in each section) were

filled with PureTemp 8 PCM and mounted. An RTD was inserted through the sealed cap into the

center of each tube. Last, the PLC control screen was used to select the desired flow rate.

Step 2: Bring tank to desired initial uniform temperature

For charging tests, the operator discharged the tank until all 18 water RTDs read ~50℉ (campus

return water temperature). For discharging tests, the operator charged the tank until all 18 water

RTDs read ~40℉ (campus chilled water temperature). For nominally augmented tests, the tank

31
was charged or discharged until all 18 water and PCM RTDs read the appropriate initial

temperature. This ensures there was no thermocline present in tank.

Step 3: Begin charging or discharging Test

Refer to Figure 5.2 for a schematic of the LABTES tank explaining which inlet and outlet valves

should be open and shut for specific charging and discharging processes. For charging tests, the

bottom manual inlet valve was opened, and the top inlet valve was closed. For discharging tests,

the top manual inlet valve was opened, and the bottom inlet valve was closed. Then the start button

on home screen of the PLC control screen was pressed to begin the test.

Step 4: Finish charging or discharging Test

The system was allowed to run until the tank reached the new initial uniform temperature. The

time when the tank’s outlet RTD temperature reached the average temperature of inlet water ±.5℉

was recorded as the charge/discharge time. The system was allowed to continuously charge or

discharge in order to set up for the next test. If a charge test was performed first, the tank continued

charging until the beginning of the next discharge test. Likewise, if a discharge test was performed

first, the tank continued discharging until the beginning of the next charge test. The operator then

returned to step 2 and proceeded to perform a test for opposite process. Once the test for the

opposite process finished, the operator pressed the stop button on home screen of PLC controller.

This was the end of the experiment

32
Figure 5.1: Procedure for Charging and Discharging the LABTES Tank

3. Begin
Charge/
Discharge
1. Prepare system process
configuration
2. Bring Tank to initial
(Set-up Tubes with PCM) Uniform Temperature

4. Finish
Charge/
Discharge
process

Figure 5.2: Equipment Set-Up for Charging and Discharging Processes

Charging Flow Path Discharging Flow Path


(Building
Return)
~𝟓𝟏℉

(Chiller)
nominally
~𝟒𝟎℉

Closed Inlet/Outlet Valve Open Inlet/Outlet Valve Water Flow Path

33
5.2 Data Analysis
Flow Rate

The flow rate was calculated using water level measurements taken at a rate of 1 measurement per

second by the magnetostrictive level meter located in the auxiliary tank. The volume of water that

enters the auxiliary tank in one second is calculated as the difference between two successive water

level readings multiplied by the area of the bottom of the tank. Then this rate is this value is

converted into units of gallons per minute. This is also converted to a mass flow rate (𝑚̇).

Tank Temperature difference (∆𝐓)

Regardless of whether the tank was augmented or un-augmented, the temperature difference was

calculated as the absolute difference between the initial bulk temperature of the water in the tank

and the average temperature of the water leaving the tank at the outlet. This temperature difference

was expressed in degrees Fahrenheit. In other words, the temperature difference was calculated as

the absolute difference between the cold temperature 𝑇𝑐 and the hot temperature 𝑇ℎ of the tank for

each test. The bulk initial temperature was calculated as the average of the temperature readings

from RTDs placed in the water equally spaced along the height of the tank at the time of start of

each charge or discharge test.

Thermocline Thickness (TT)

Thermocline thickness was calculated using the method suggested by Stewart (9). An RTD in the

water at about mid-height in the tank was chosen. As the thermocline rose or fell, the temperature

history of this RTD showed the thermocline. The time that the RTD measured 𝑇𝑐 +.5℉ (𝑡1 ) and the

time that the RTD measured 𝑇ℎ −.5℉ (𝑡2 ) were recorded, and the difference in minutes between

34
Figure 5.3 How to Identify the Thermocline Using the Temperature History of an RTD

35
these times was multiplied by the rise rate (RR) of water through the tank. The units of the rise

rate are inches per minute. This produced the thermocline thickness expressed in inches.

Experimental and Theoretical Charge Time

The experimental charge time was calculated as the time between when the charge or discharge

test began and the time the outlet temperature of the tank reached within .5℉ of the average inlet

temperature, expressed in minutes. The theoretical charge time was calculated by dividing the

volume of water in the tank in the gallons by the volumetric flow rate of water through the tank.

Half Figure of Merit (FOM)

The half figure of merit was calculated individually for each charge and discharge process. It was

calculated as the integrated capacity (𝐶𝑖𝑛𝑡 ) of the storage tank divided by the maximum capacity

possible (𝐶𝑚𝑎𝑥 ) for each test for the same time duration. The integrated capacity of the tank is

calculated as follows where 𝑐 is the specific heat of water, 𝑇𝑖𝑛 is the average inlet temperature of

the tank, 𝑇𝑜𝑢𝑡 is the average inlet temperature of the tank.

𝑡 ̇ (Eq. 5.1)
𝐶𝑖𝑛𝑡 = 𝑚̇ 𝑐 ∑𝑡𝑓𝑖𝑛𝑎𝑙 |𝑇 − 𝑇𝑜𝑢𝑡 | ∆𝑡
𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑖𝑛
where ∆𝑡 = 1 second

𝑡 ̇ (Eq. 5.2)
1 𝐶𝑖𝑛𝑡 𝑚̇ 𝑐 ∑𝑡𝑓𝑖𝑛𝑎𝑙 |𝑇 − 𝑇𝑜𝑢𝑡 | ∆𝑡
𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑖𝑛
𝐹𝑂𝑀 = ∙ 100% = ∙ 100%
2 𝐶𝑚𝑎𝑥 𝑚̇ 𝑐 (𝑇ℎ − 𝑇𝑐 )(𝑡𝑓𝑖𝑛𝑎𝑙 − 𝑡𝑖𝑛𝑖𝑡𝑖𝑎𝑙 )

Note that the temperature difference |𝑇𝑖𝑛 − 𝑇𝑜𝑢𝑡 | is an absolute value in this equation because in

discharging, 𝑇𝑜𝑢𝑡 is less than 𝑇𝑖𝑛 , and in charging, the 𝑇𝑖𝑛 is less than 𝑇𝑜𝑢𝑡 . 𝐶𝑚𝑎𝑥 increases faster

than 𝐶𝑖𝑛𝑡 once |𝑇𝑖𝑛 − 𝑇𝑜𝑢𝑡 | becomes small after the water in the tank has been replaced once.

36
System Efficiency

The system efficiency (𝜂) was calculated as the ratio of the integrated thermal capacity stored in

the tank during charging (𝐶𝑖𝑛𝑡 𝐶 ) to the integrated thermal capacity retrieved from the tank during

discharging (𝐶𝑖𝑛𝑡 𝐷 ).

𝐶𝑖𝑛𝑡 𝐷 (Eq. 5.3)


𝜂= 𝐶𝑖𝑛𝑡 𝐶

37
6 Sensible (Un-augmented) Stratified TES Tank

Before augmenting the LabTES tank with pipes full of PCM, it was important to first verify that

the tank functioned as expected without any augmentation. Therefore, as a control for the

augmented stratified tank testing, the tank was first tested without any heat exchanger pipes in the

tank. The PVC rack with 18 RTDs placed linearly along its height was suspended in the center of

the tank. This was called the sensible configuration because the system did not involve any latent

heat storage. The flowrate of water through the tank has important effects on the charge/discharge

time and the thermocline and thus on the efficiency of the tank. It was found that the precision of

the flowrate was difficult to control. Therefore, many separate charge and discharge cycles were

run in the sensible configuration using varying flowrates between 0 and 2.6 GPM to characterize

the effects of flowrate on tank behavior. Results of these tests are listed in Tables 6.1 and 6.2.

Table 6.1: Results for Sensible Charging Tests

Test Average ∆T Experiment Theory Experiment Integrated System Half FOM


Flow (℉) Charge Charge 80% TT Capacity Efficiency (%)
# Rate Time Time (in.) Charging 𝜂
(GPM) (min.) (min.) (Ton-Hrs) (%)

1 0.86 9.17 264 187 12.13 1.28 81.0% 88.5%

2 0.89 7.86 245 182 12.09 1.10 81.3% 92.5%

3 1.47 10.20 166 109 17.17 1.45 80.8% 83.3%

4 1.57 5.91 129 102 9.87 0.83 82.0% 99.1%

5 1.83 11.75 105 88 9.18 1.46 80.5% 92.7%

6 2.14 9.89 97 75 11.30 1.33 80.8% 92.8%

7 2.21 9.60 107 73 14.85 1.36 80.9% 85.7%

8 2.25 10.26 98 72 11.43 1.39 80.8% 88.3%

38
Table 6.2: Results for Sensible Discharging Tests

Test Average ∆T Experiment Theory Experiment Integrated System Half


Flow (℉) Discharge Discharge 80% TT Capacity Efficiency FOM
# Rate Time Time (in.) Discharging 𝜂 (%)
(GPM) (min.) (min.) (Ton-Hrs) (%)

9 0.39 10.77 412 412 11.07 1.07 80.7% 88.1%

10 0.95 9.22 180 170 8.63 0.99 81.0% 90.4%

11 1.04 8.93 169 155 7.67 1.01 81.0% 92.5%

12 1.59 10.94 106 102 6.83 1.16 80.6% 90.5%

13 1.73 10.49 116 93 9.24 1.25 80.7% 85.2%

14 2.15 9.90 90 75 12.08 1.03 80.8% 76.9%

15 2.46 10.30 77 65 10.29 1.19 80.7% 87.5%

6.1 Effects of Varying Flow Rate

According to the literature, varying flowrate has several effects on the behavior of the tank.

Increasing flow rate increases the Reynolds number and the amount of mixing within the tank,

thus decreasing efficiency (Half FOM). However, because the flowrates used in this experiment

do not cause the Fr number to be greater than 1, the mild increase in flowrate examined in this

experiment should not significantly affect the efficiency. It may also be observed that increasing

flow rate did not directly correlate to increased thermocline thickness in the above tests. Rather, it

is theorized that the residence time of the water in the TES tank in conjunction with the heat gain

through the sides and top of the tank from the surrounding air that varied in temperature from

approximately 65 to 70℉ caused the thermocline thicknesses to be much less predictable than

expected.

39
6.2 Charge Time
Figure 6.1 shows the outlet water temperatures during charging and discharging (Tests # 6 and #

12). The charge time (𝑡𝑐 ) for each test is marked by a dotted line. As shown in Figure 6.2,

experimental discharge times closely follow the theoretical model for charge time at a given flow

rates. Values for experimental charge time are consistently higher than theoretical values. This

may be because when charging the tank, the inlet water must work against ambient heat transfer

warming the tank whereas in discharging the ambient heat transfer works with the warmer inlet

water to bring the tank back up to the temperature of the return water. For both charging and

discharging tests, it was found that agreement with the theoretical model was best at higher flow

rates. This is most likely because increasing flow rate causes proportional decrease in

charge/discharge time allowing less time for unwanted external heat gain to occur.

6.3 Thermocline and Thermocline Thickness


Figure 6.3 shows the passage of the thermocline with time for three different charging tests (Tests

# 1, 3, and 7). Figure 6.4 shows the passage of the thermocline with time for three different

discharging tests (Tests # 10, 12, and 15). Each subplot shows one thermocline profile for each of

eight RTDs located along the vertical axis of the TES tank. This method of using the temperature

profile of a single RTD to view the thermocline profile is described in detail by Dr. William

Stewart [9] and echoed in studies by Dr. William Bahnfleth. During charging tests, the thermocline

passes the lowest RTD in the tank first and the highest RTD in the tank last, and for discharging

tests, the thermocline passes the highest RTD in the tank first, moving downward.

In Figures 6.3 and 6.4, it is obvious that increasing flowrate causes a proportional decrease in

charge time. The thickness and shape of the thermocline, however, is not accurately depicted in

40
Figure 6.1 Water Outlet Temperatures Vs. Water Inlet Temperatures

𝒕𝒄 =97 min.

Sensible Discharge𝒕 Flow Rate: 1.59


=106 min.
𝒄

GPM Delta T: 10.94℉

Figure 6.2 Relationship Between Charge/Discharge Time and Flow Rate

𝐓𝐚𝐧𝐤 𝐕𝐨𝐥𝐮𝐦𝐞
𝐓𝐡𝐞𝐨𝐫𝐞𝐭𝐢𝐜𝐚𝐥 𝐂𝐡𝐚𝐫𝐠𝐞(𝐃𝐢𝐬𝐜𝐡𝐚𝐫𝐠𝐞) 𝐓𝐢𝐦𝐞 =
𝐅𝐥𝐨𝐰 𝐑𝐚𝐭𝐞

41
Figure 6.3 Thermoclines Measured from RTDs at Different Heights in the Tank

Figure 6.4 Thermoclines Measured from RTDs at Different Heights in the Tank

42
these figures because the number of minutes the thermocline takes to pass is not completely

indicative of thermocline thickness. Note that for lower flow rates, the thermocline profiles seem

more stretched out over the time axis. This does not mean that thermocline thickness is greater at

lower flow rates. This time it takes for the thermocline to pass must be multiplied by the average

rise rate in the tank in order to calculate the experimental thermocline thickness values listed in

Tables 6.1 and 6.2. And the average of these 8 thermocline thickness values from the different

RTDs is given in the 6th column of Tables 6.1 and 6.2.

From beginning to end of the charging or discharging process, the thermocline thickness does not

remain constant as the thermocline traverses the tank. Figure 6.5, shows the thermocline thickness

measured at each RTD along the height of the tank for discharging Test #15 and charging Test #

6, demonstrating how the thermocline thickness behaves over the course of the thermocline

passing from top to bottom or from bottom to top of the tank. Just as Bahnfleth observed in his

own studies, for each of Tests # 1 through # 15, the thermocline thickness increased slightly as the

thermocline traversed the tank from inlet to outlet. This increase in thickness is thought to be

attributed to unwanted mixing and diffusion across the thermocline. Longer residence times of

water in the tank contribute heavily to this diffusion.

Average thermocline thickness is shown in Figure 6.6 for all sensible tests. The effects of ∆T and

flow were examined to determine their possible influence on thermocline thickness. Figure 6.6

plots thermocline thicknesses with respect to these factors. By this analysis, it was concluded that

∆T, experimental charge/discharge time, and heat losses effect thermocline thickness more than

increase in flow rate provided that Froude number is less than 1. On average, in these tests, the

thermocline Thickness is approximately 12” thick and occupied 17% of the height of the TES tank.

43
Figure 6.5: Showing Thickening of Thermocline during Charging & Discharging

Water RTDs 2,3,4,6,7,8


Water RTDs 2,3,4,6,7,8 Test # 6

Test # 15

Figure 6.6: Relating Thermocline Thickness to Temperature Difference and Flow Rate

44
In studies by Bahnfleth on larger commercial size TES tanks, the thermocline took up only 10 %

of the tank [10]. This suggests that the LabTES tank should have been designed to be taller.

6.4 Integrated Capacity and System Efficiency


Figure 6.7 shows the values of integrated capacity for Tests # 1 to # 15. (7th column of Tables 6.1

and 6.2) Note that integrated capacity increases as ∆T increases. The average integrated capacity

for values of ∆T=8 to 12 ℉ was 1.37 Ton-hours in charging and 1.10 Ton-hours in discharging.

This trend agrees well with the mathematical definition of integrated thermal capacity given earlier

in Chapter 5. Discrepancies between the trend lines and the data point may be attributed to varying

conditions in the overall cooling system the tank is embedded in as well as measurement error in

the charge/discharge time and flow rate. Measured integrated capacity values needed to charge the

tank are consistently higher in charging than in discharging. This is due to unwanted heat transfer

between the tank and the environment. This contributes to an overall system efficiency and half

FOM of less than 100% (to be shown next).

Note none of the charging tests have a corresponding discharge test with the same exact ∆T.

Indeed, most commercial stratified TES tanks do not run discharge processes immediately

following charging processes but instead store the built up thermal capacity for some time before

retrieving it. In order to better characterize the values of integrated capacity over the same range

of ∆T, one linear trend line was drawn for the charging data points and one trend line for the

discharging data points because by definition, capacity should be linearly related to ∆T. The values

of integrated capacity on these two trend lines are then used to calculate the system efficiency

values (8th column of Tables 6.1 and 6.2) for each ∆T as plotted in Figure 6.8. The system

efficiency compares the integrated capacity required to charge the tank to the integrated capacity

45
Figure 6.7 Relationship Between Integrated Capacity and Temperature Difference

Figure 6.8 Linear Estimation of System Efficiency

46
retrieved from the tank during discharging. According to calculations, on average 80.9% of the

capacity stored in charging is retrieved for tests with values of ∆𝑇 = 6 to 12 ℉ with slightly higher

values of system efficiency observed at lower temperature differences.

6.5 Half FOM


The last column of Tables 6.1 and 6.2 show the average half FOM for charging tests with a ∆T

between 9 and 11 ℉ is 88.6%, and the average half FOM for discharging tests with a ∆T between

9 and 12 ℉ is 86.4%. Half FOM is calculated as per Eq. 5.2. The initial bulk temperature in the

tank and the inlet water temperature varied for each test causing varying values of ∆T. Tests with

values of ∆T outside these ranges were not included in these averages because tests with smaller

values of ∆T tend to exhibit significantly higher values of half FOM. This trend can be seen in the

results for Tests # 4 and # 11. Figure 6.9 shows data for half FOM plotted in relation to ∆T and

flow rate. Though linear trend lines are shown, note that the data points are scattered, suggesting

no obvious function relating these terms. Analysis of the data showed that the calculation of half

FOM is not a strong function of m ̇ or ∆T. Because the Froude number for all tests is less than 1,

it was not expected that half FOM would be significantly affected by mass flow rate (𝑚̇). This

assumption agrees with the data plotted.

Half FOM may potentially depend on several factors within the test. For this reason, further

analysis was done in order to determine the most influential factors effecting half FOM. Figure

6.10 shows the relationship between thermocline thickness and half FOM for tests with similar

values of ∆T. Only tests with values of ∆T between 9 and 12℉ are shown. Charging and

discharging tests are shown with separate trend lines, and both trends show a steady decrease in

thermocline thickness as half FOM improves. In this respect, the LabTES tank performed as

expected.

47
Figure 6.9 Relating Half FOM to Temperature Difference and Flow Rate

Figure 6.10 Relationship Between Thermocline Thickness and Half FOM

Average Thermocline Thickness Vs. Half FOM

As TT increases, Half FOM decreases

48
As demonstrated in Figure 6.10, experimental values of Half FOM do depend inversely on

thermocline thickness. This conclusion agrees with similar findings in past literature. However, it

was found that charge/discharge time most effected half FOM. Half FOM dropped if the tank

was charged or discharged for much longer than it takes for the mass of water in the tank to be

replaced once.

6.6 Conclusions

Based on the data shown, the LabTES tank performed as expected in accordance with result from

prior literature. Relationships between experimental flow rates and plug flow charge/discharge

times and the relationships between experimental thermocline thickness and half FOM are

consistent with those observed in other TES tanks studied by Bahnfleth and Stewart. The

thermocline (~12” thick) in LabTES tank was easily and repeatably established and moved up and

down to charge or discharge the chilled water tank over the range of flow rates and operating

temperature differences investigated. Larger than expected time increase were attributed to

thermocline thickness and external heat gain. The successful design, building, and sensible test

data obtained from the stratified LabTES tank produced the necessary reference data/ comparison

for the augmented LabTES tank testing.

The significant heat gain due to the surface of the water being exposed to ambient temperature and

the limited amount of insulation on the sides and bottom of the tank prevent the LabTES tank from

keeping chilled water cold for more than a day. Heat gain in the LabTES tank may also be due to

the smaller size of the tank in comparison to the much larger TES tanks used on campus. For this

reason, the LabTES tank cannot adequately store thermal energy in accordance with the schedule

of energy output for the power plant. Heat gain over extended charge/discharge times was

considered during further testing. To minimize heat gain in augmented tests, charge/discharge

49
times were reduced by using higher flowrates that would still allow for a Froude number less than

1. Adding more outer insulation to tank would prevent unwanted external heat gain and improve

FOM but was not done to keep the tank parameters the same, except for the inclusion of PCM, to

keep the comparison fair. External heat gain is less of a problem in the larger TES tanks at

University of Cincinnati, which can maintain water at a chilled temperature for at least 7 days

because they hold a much larger mass of water. The large TES tanks are also underground,

providing excellent insulation from ambient temperatures.

50
7 Nominally Augmented Stratified TES Tank

Following system validation (i.e. sensible testing), all further tests were conducted with a

combination of PCM-filled tubes placed along the PVC rack to augment the thermal storage

capacity of the tank. See Figure 7.1. A combination of nine copper tubes was chosen as the nominal

case for augmented testing. Copper was chosen as the pipe material for its high thermal

conductivity. One RTD was placed in each PCM tube and one in the water just outside each tube,

making nine PCM RTDs and nine water RTDs. As in sensible tests, the tank was first operated

using varying flowrates between 0 and 2.6 GPM. Further tests were performed with a set point

flow rate of 2.5 GPM. The results of these tests are listed in Tables 7.1 and 7.2. Experimental

charge/discharge time was examined in two different ways. The experimental charge/discharge

time for the water in the tank is defined as the time from starting the charge/discharge to when the

top/bottom water RTD reaches within .5 ℉ of the inlet water temperature and is recorded on the

right side of the Experimental Charge Time column. On the left side is the time that it took for the

top/bottom PCM RTD to reach the same criteria. The integrated capacity column shows values

calculated based on experimental charge/discharge time of the PCM on the left and of the water

on the right.

51
Figure 7.1 Schematic of LabTES Tank

Tank is augmented with nominal configuration of copper pipes, showing numbered copper
tubes and positions of RTDs in both water and PCM.

Tube 1

Tube 2

Tube 3

Tube 4
Tube 5
Tube 6
Tube 7
Tube 8
Tube 9

Water RTDs PCM RTDs

52
Table 7.1: Results for Augmented Charging Tests
Test Average ∆T Experiment Theory Experiment Integrated System
Flow Rate (℉) Charge Time Charge 80% TT Capacity Efficiency
# (GPM) (min.) Time (in.) Charging 𝜂
(min.) (Ton-Hrs) (%)
PCM Water PCM Water
16 0.76 8.87 1062 357 208 10.73 1.50 1.47 86.6%

17 1.28 10.95 500 152 123 9.72 1.98 1.48 89.1%

18 2.12 9.54 108 91 74 9.52 1.23 1.20 87.5%

19 2.18 9.31 107 90 72 9.41 1.27 1.24 87.2%

20 2.20 8.79 107 88 72 9.23 1.16 1.13 86.5%

21 2.22 8.71 108 91 71 10.72 1.19 1.17 86.3%

22 2.22 8.10 104 89 71 11.88 1.11 1.08 85.3%

Table 7.2: Results for Augmented Discharging Tests


Test Average ∆T Experiment Theory Experiment Integrated System
Flow Rate (℉) Discharge Discharge 80% TT Capacity Efficiency
# (GPM) Time Time (in.) Discharging 𝜼
(min.) (min.) (Ton-Hrs) (%)
PCM Water PCM Water
23 1.65 10.94 271 144 95 14.08 1.31 1.26 89.1%

24 2.44 12.09 170 110 65 16.90 1.51 1.46 90.2%

25 2.48 9.06 405 98 63 19.34 1.39 1.06 86.8%

26 2.51 9.94 317 106 63 24.79 1.36 1.19 88.0%

27 2.63 9.79 283 86 60 14.43 1.33 1.20 87.8%

53
7.1 Effects of Subcooling and Latent Heat
Figure 7.2 shows the behavior of the temperature in each tube of PCM with respect to time during

the charging process in Test # 21. Each tube exhibits the same pattern of behavior. The charging

process is broken down here into four different regions delineated by the letters listed on Figures

7.2 and 7.3. The average inlet chilled water supply temperature in Figure 7.2 is 41.2.℉. For the

sake of explanation, the letters follow the temperature history of Tube 1.

A-B: Sensible Heat Transfer- As chilled water enters the tank, the temperature of the PCM inside

the tubes decreases. The nine tubes are placed one above the other inside the tank. Therefore, the

PCM temperatures decrease one after another as the chilled water rises from section 9 at the bottom

of the tank to section 1 at the top of the tank until all tubes reach the fusion temperature of the

PCM.

B-C: Subcooling- The temperature of the PCM then drops significantly below the expected fusion

temperature of 46.4 ℉ (8℃) during subcooling. Puretemp8 does not solidify immediately upon

reaching the fusion temperature but instead must be cooled to an even lower temperature in order

for crystallization to begin so that freezing can occur. This phenomenon is referred to as

subcooling. The average temperature that the PCM in each tube must be cooled to, before

crystallization occurs, is 41.9 ℉ for the data as shown in Figure 7.2.

C-D: Latent Heat Transfer- After subcooling, all nine PCM temperatures then rise to about 44

℉ first the bottom tube and last the tubes near the top. During latent heat transfer, the temperature

of the PCM stays fairly constant as the PCM releases its latent energy of phase change.

D-E: Sensible Heat Transfer- Once the latent energy of phase change is transferred, it can be

assumed that all of the PCM is in a solid state. From this point forward, the frozen PCM cools

54
sensibly toward the temperature of the inlet water. Note that the rate of temperature drop after

phase change is not as sharp as compared to the sensible drop before the phase change due to

much smaller ∆T=TFusion-Tinlet≈3℉.

Variances in the amount of subcooling observed in each tube may be due to small variances in

sample size of PCM in each tube and the random nature of crystallization. Each individual liquid

element in a PCM sample reaches the temperature just below the freezing point crystallizes at a

slightly different time. Nucleation sites for crystallization are formed, and the rate of heat transfer

into the PCM tube as well as the rate of heat transfer between individual liquid PCM elements then

determines the rate of crystallization and thus the amount of subcooling. The effects of subcooling

are especially significant when the temperature differential between the inlet water temperature

and the freezing point is small. The larger the degree of subcooling is in comparison to this

temperature difference, the more negative impact subcooling has on the efficiency of the system.

Figure 7.3 shows the behavior of the temperature in each tube of PCM with respect to time during

discharging in Test # 26 The average inlet return water temperature is 51.8.℉.

A-B: Sensible Heat Transfer- At the onset of discharging, the tubes of PCM begin quickly rising

in temperature in order, one after the next, starting with tube 1 at the top of the tank where the

return water enters and moving downward to tube 9 at the bottom of the tank. However, because

subcooling doesn’t occur during the melting process, points B and C are at the same time.

C-D: Latent Heat Transfer- Once the tubes reach the fusion temperature and melting begins, the

temperature rises much more slowly, remaining fairly constant, until the latent heat energy is

overcome. Like subcooling, melting is much less predictable than sensible heating, and the tube

55
Figure 7.2 Tube Temperatures and Phase Change During Charging
Sensible Cooling

Tubes from bottom of


tank to top. Bottom
tubes change first.

Subcooling
B
Latent Heat Transfer
C D Sensible Cooling

Figure 7.3 Tube Temperatures and Phase Change During Discharging

Sensible Heating
E

Latent Heat Transfer

Sensible Heating
B,C D
A

56
temperatures no longer increase in the same order as they did before melting because each one

went through randomized intermolecular reactions throughout the melting process.

D-E: Sensible Heat Transfer- After melting, PCM temperature increases sharply once the latent

heat energy is overcome, moving toward the warm inlet water temperature.

Note that in discharging, the difference between the PCM fusion temperature and temperature of

the return water entering the tank is about 6 ℉, whereas in charging the difference between the

PCM fusion temperature and the chilled water entering the tank is only about 3 ℉. Because this

temperature difference is larger in discharging, the latent heat of phase change is overcome more

quickly than in charging. This can be seen when comparing the time scales of Figures 7.2 and 7.3.

7.2 Charge/Discharge Time

Figure 7.4 shows the charge and discharge times for both the water and the PCM in the tank during

the augmented charge and discharge Tests, #16 through # 27. The charge/discharge times for water

in the nominally augmented tests were comparable to those of the sensible tests. As in sensible

tests, the experimental charge/discharge times for water are longer than the expected values.

However, the discrepancy between experimental and theoretical discharge times for water did

increase slightly in the augmented case. Charge time for both PCM and water decreases as flow

rate increases just as it did in sensible tests. The PCM tubes take up volume, leaving less water in

the tank in the augmented case. Therefore, values of theoretical charge time (plotted by the black

line in Figure 7.4) in augmented tests are slightly less than those in sensible tests at similar flow

rates. Experimental charge/discharge times in Tables 7.1 and 7.2 show the PCM took significantly

longer than the water to approach the inlet water temperature in both charging and discharging.

This is clearly shown by the PCM temperatures for discharging in Figure 7.5. This is because of

57
Figure 7.4 Experimental Charge Times of Water and PCM

Figure 7.5 Comparing PCM and Water Charge Times

(Top) Time scale zoomed in (Bottom) Time scale zoomed out

58
the latent heat involved in the phase change and added conduction resistance of the PCM itself.

Note that for charging at flowrates less than 2 GPM the difference between PCM and water charge

times is significantly higher. The increased charge times for PCM indicate that in order to take full

advantage of the latent and sensible thermal storage of the PCM, the tank must be charged for

longer than it takes for the mass of water in the tank to be replaced once. The increased discharge

time means that it took longer for the water exiting the tank to heat up to the return water

temperature, implying that the tank was able to output cooler water for a longer period of time

during discharging.

7.3 Thermocline Thickness

Using the elapsed time of the thermocline passing and the rise rate, a value of thermocline

thickness was calculated using each of the nine water RTDs. The most accurate thermocline values

came from the temperature histories of the water RTDs located in the middle sections of the tank,

farther from the distribution ring openings. The average of these more accurate values was

recorded as the experimental 80% thermocline thickness for each test in Tables 7.1 and 7.2. Figures

7.6a and 7.6b, show the temperature traces for Tests # 22 and # 26, respectively, as compared to

the thermoclines of these augmented tests to those of sensible Tests # 7 and # 15, which exhibited

similar flow rates. When comparing the results for thermocline thickness in the nominal

augmented case to those of the sensible case listed in Tables 6.1 and 6.2, it can be concluded that

the presence of PCM has the most effect on system performance during discharging. In the

augmented case, the thermocline thickness in discharging is significantly larger. Because the PCM

changes phase early in the discharge process, when the difference between the initial bulk water

temperature in the tank and the inlet water temperature is higher, the PCM changes phase more

59
Figure 7.6a Thermoclines Measured from RTDs at Different Heights in the Tank

Figure 7.6b Thermoclines Measured from RTDs at Different Heights in the Tank

60
Figure 7.7: Relating Thermocline Thickness to Temperature Difference and Flow Rate

61
quickly. When the PCM changes phase more quickly, the faster internal heat transfer going on in

the tank increases thermocline thickness. Conversely, the PCM changes phase late in the charging

process when the temperature difference is small, and therefore phase change is slow and effects

thermocline thickness much less. In charging, the values of thermocline thickness for the

augmented case are not significantly different from those of the sensible case.

Figure 7.7 plots thermocline thickness with respect to temperature difference and flow rate for

both sensible and augmented tests. As in sensible tests, thermocline thickness in augmented tests

was not a strong function of temperature difference or flowrate for Froude numbers less than 1.

Note that with the small packing factor used in these augmented tests, the thermocline and thermal

stratification of water in the tank was not disrupted. Increasing packing factor, though increasing

thermal capacity, if made too large is expected to disrupt stratification, making the tank ineffective.

7.4 Thermal Capacity


Figure 7.8 (top) shows the temperature histories over time of the inlet and outlet RTDs for the

discharge Tests # 12 (sensible) and # 23(augmented with PCM). The area enclosed between the

sensible inlet and outlet lines is equal to the thermal capacity for cooling released to the power

plant system during the sensible (no PCM) discharge process. The area enclosed between the

nominal inlet and outlet lines is equal to the thermal capacity for cooling released to the power

plant system during the nominally augmented (with PCM) discharge process. These areas are

mathematically expressed as integrals from the starting time of the discharge process to the time

that the inlet line intersects with the outlet line (within .5 ℉). Note that this intersection occurs

much later in the nominally augmented case than it does in the sensible case. When calculating the

area from the measured data in 1-minute time increments, the cooling capacity delivered by the

sensible discharge is 1.16 ton-hours, whereas the cooling capacity delivered by the nominally

62
Figure 7.8 Comparing Sensible and Nominal Outlet Temperatures

With PCM

With PCM

63
Figure 7.9 Comparing Integrated Capacity in Charging and Discharging

Figure 7.10 Linear Estimation of System Efficiency

64
augmented discharge is 1.31 ton-hours. From this it can be concluded that if the PCM is fully

charged and fully discharged, the nominal augmentation of the TES tank does provide significantly

more cooling capacity to the power plant system during discharging.

Figure 7.8 (bottom) shows the temperature histories over time of the inlet and outlet RTDs for the

charge tests # 6 and # 18 (without and with PCM, respectively). Once again, the thermal capacity

for both the sensible and augmented tests can be calculated as the area enclosed between the

respective inlet and outlet lines. When calculated, the cooling capacity needed, based on water

charge time, for the sensible charge is 1.33 ton-hours, and the cooling capacity required by the

nominally augmented charge is 1.20 ton-hours, .13 ton-hours less since the PCM tubes take up

volume, leaving less water in the tank in the augmented case.

Figure 7.9 plots the integrated capacity, based on water charge time, for nominally augmented tests

16 to 27 along with linear trend lines to estimate the relationship between integrated capacity and

temperature difference. These linear approximations of integrated capacity in charging and

discharging are then used to calculate the values of system efficiency plotted in Figure 7.10. The

average system efficiency for nominally augmented tests was 87.5%, significantly higher than in

sensible tests due to higher integrated capacity in discharge. Note that the system efficiency is

slightly higher when temperature difference is higher. This is because larger temperature

differences allow for faster phase change of the PCM, storing or releasing large amounts of thermal

energy with less opportunity for external heat gain.

Notice that the values of system efficiency in discharging for the augmented case is significantly

higher than that in the sensible discharging case. The higher system efficiency in the augmented

discharging case can be attributed to a significant discrepancy in the average mass of water that

went through the tank during augmented charging tests and average mass of water that went

65
through the tank during augmented discharging tests. In Tables 7.1 and 7.2, observe that the values

for flow rate in discharging are consistently higher than those in charging. This may also possibly

contribute to the higher values of thermocline thickness observed in discharging. Also, the

experimental water charge time is generally much greater in discharging than it is in charging.

This means that over all a significantly greater amount of water mass traveled through the tank

during discharging tests than in charging tests. This means that values for integrated capacity in

discharging were greater than they would have been had the discharging tests used the same

amount of water mass as was used in the charging tests. These higher values of integrated capacity

in discharging artificially boosted values of system efficiency. It can be concluded that both system

efficiency and integrated capacity are not exclusively related to temperature differential but also

to the mass the passes through the tank. Therefore, in future tests, the water mass used in charging

and discharging tests should be kept constant.

Figure 7.11 compares the integrated capacity retrieved from the TES tank during discharging from

the sensible case verses the nominally augmented case. Note that both trend lines for the nominally

augmented case show greater integrated capacity values than the trend line for the sensible case.

The trend line for the nominally augmented integrated capacity, based on PCM charge time, shows

the greatest augmentation because this is the case in which the tank is discharged for the longest

period of time. Further tests are desired in order to verify the slopes of these trend lines. However,

the trend lines, plotted in Figure 7.11 were used to calculate the percentage of augmentation shown

in Figure 7.12 by dividing the difference between the integrated capacity given by either nominal

trend line and the sensible trend line at the same value of ∆𝑇 in Figure 7.11 over the value of

integrated capacity given by the sensible trend line and multiplying that by 100%. Based on this

66
Figure 7.11: Comparing Integrated Capacity in Discharging

Figure 7.12 Percent Augmentation Over Sensible Capacity

67
calculation, it was found that when discharging the nominally augmented tank only long enough

for all RTDs in the tank to reach near the inlet temperature (i.e. curve based on water charge time),

the integrated capacity retrieved from the augmented tank was 5-15% higher than the integrated

capacity retrieved from a completely sensible tank operated at the same temperature difference.

Measured augmentation based on PCM charge time is much higher, even higher than our ideal

theoretical prediction obtained in chapter 3 of 9.821% at ∆T= 6 ℃ (10.8 ℉). However, the

measured augmentation calculation gives more reasonable values as the temperature difference is

increased and becomes on par with the theoretical predictions at ∆T= 12 ℉.

7.5 Conclusions
Tests # 16 through # 27 show that the TES tank still performs its original function when

augmented. In fact, it was found that the augmented TES tank outputs a larger mass of cooler water

over a longer period of time than the purely sensible tank. The practicality of augmenting the

thermal storage of a stratified tank using tubes full of PCM can perhaps be measured by the

increased discharge times of the water in the tank rather than by increased half FOM or system

efficiency. Increased water discharge time indicates that the water exiting the tank during

discharging takes longer to warm up. This means that at a constant mass flow rate, the augmented

tank delivers a larger mass of cool water to the system. If given this, then augmenting the tank

with PCM delivers output chilled water for longer periods of time, and hence more capacity. This

would be an advantage to any system that the TES tank is employed in.

One must, however, also consider the amount of cooling capacity used to charge the TES tank in

the first place as well as the amount of cold storage lost to the tank’s surroundings during the time

in between the end of the charging process and the start of the discharge process. If the tank is not

68
well insulated and operated in a specific manner, this increased capacity may not necessarily

improve system performance.

Based on these nominally augmented experiments, the best practices for charging the LabTES tank

are as follows. The augmented tank must be charged until the PCM reaches the inlet water

temperature, ensuring that the PCM is in the solid phase. When using flowrates above 2 GPM, this

does not take a significantly longer time than it would take for the water to reach the inlet water

temperature. Charge time for the PCM becomes significantly longer than charge time for water

when using flowrates below 2 GPM, decreasing half FOM. For this reason, all further experiments

used flowrates greater than 2 GPM.

The most efficient way to discharge the TES tank is to begin discharging immediately at the end

of the charging process and discharge the tank completely. However, most commercial TES tanks

are not operated that way and are instead discharged only enough to cover a specific cooling load

needed by the system at the time. According to Figure 7.4, the largest difference between the

nominal and sensible discharging capacities occurred at the end on the discharge process.

Therefore, in order to take full advantage of the extra capacity provided by augmentation, a TES

tank must be discharged completely.

Ideally, we would want the PCM to charge in the same amount of time it takes for the thermocline

to pass through the tank. However, as the experiments here revealed, because the PCM takes longer

to charge and discharge than it takes for the tank water to be replaced, the overall efficiency of the

charging process would have to be sacrificed in order to achieve the maximum possible capacity.

It is hypothesized that increasing the rate of heat transfer between the PCM and the water might

allow for the PCM to charge in a shorter time period. However, in order for the tank to gain more

69
capacity the tank would still have to be charged for longer than it takes to replace the mass of water

in the tank once.

The packing factor is the percentage by mass of the storage medium in the tank that is PCM, and

the mass packing factor for the nominal augmented case is only 1.63%. The rest of the storage

medium in the tank is water. Because the packing factor is small, it was not expected that the

presence of PCM in the tank would affect the thermocline significantly. While an increased

packing factor is desirable to increase latent storage capacity, increasing the packing factor may

have a more significant negative effect on the thermocline and the system efficiency for the tank.

70
8 Overall Conclusions and Future Work

Over all, the LabTES experiments were successful and provided a more detailed knowledge of

challenges involved in the implementation of stratified cold storage augmented with PCM. The

following is a summary of the conclusions and lessons learned from these experiments.

8.1 Conclusions
All original objectives of this study were met. The nominally augmented tank outputs a larger mass

of cooler water over a longer period of time than the purely sensible tank. According to the

measured performance shown in Chapter 7, the nominally augmented tank delivered

approximately 10% more thermal capacity during the discharging process than the sensible tank

delivered, making the experiment an overall success. Regardless of the internal heat transfer

involved with the PCM augmentation, the laboratory TES tank still performed its original function,

and the thermocline was not significantly disrupted. These findings provided proof of concept for

further research in integrating PCM with stratified tanks.

The laboratory stratified TES tank was designed and was installed with its accessories and

instrumentation in line with the chilled water and return water supplies of the University of

Cincinnati east power plant. The tank containing only water and using only sensible thermal

storage was charged and discharged multiple times in order to gather reference data. Using this

data, it was verified that the laboratory TES tank behaved as expected in accordance with literature

describing other commercial TES tanks. However, the small size and depth of the laboratory TES

tank did allow for increased unwanted external heat transfer and a shorter residence time for the

thermocline to form and travel the height of the tank. The tank was successfully augmented with

a PVC rack holding 9 copper tubes filled with PureTemp8 PCM. Charge and discharge tests were

71
run in order to obtain new tank performance data which was then compared to the reference data

from the sensible tests.

Current set-up requires optimization. In future studies, there are several improvements that could

be made to the laboratory TES tank system. It is recommended that the laboratory TES tank depth

be increased to at least 15 ft. There should also be a significant increase in the amount of insulation

around the tank. And the air temperature outside the tank should be carefully controlled and made

as constant as possible. It is also essential to improve the flow rate control and measurement

instrumentation. An increased number of RTDs to measure temperature both inside the PCM tubes

and in the water surrounding them inside the tank would offer better insight into the behavior of

the thermocline and allow for more accurate thermal modeling of the tank.

The performance enhancement for tank when using PCM was determined using calculations of

integrated capacity, half FOM, and system efficiency as described by Bahnfleth in his experiments

with full scale stratified tanks [6]. Because the nominally augmented tank provided more increased

integrated capacity in discharging, the values of system efficiency in the nominally augmented

tests were also higher than in sensible tests. However, it is desirable to have more data points to

verify that this level of augmentation can be consistently reproduced. Possible negative effects on

thermocline and tank stratification were considered. It was theorized that increasing the packing

factor of the PCM would eventually degrade the thermocline. However, the small mass packing

factor used in this study did not prevent stratification. Two extra discharge cycles were ran using

different numbers of PCM tubes in the tank to examine the effect of larger and smaller packing

factors on the thermocline. This data can be viewed in Appendix A, Figure A.6.

72
This study ultimately provided proof of concept for the inclusion of PCM to augment stratified

TES tanks at University of Cincinnati. A system of cheaply produced cylindrical heat exchanger

tubes could feasibly be used to integrate latent thermal storage with the sensible thermal storage

of the stratified TES tanks. However, further testing will be required in order for full integration

to take place.

8.2 Recommended Future Work


Ideally, it would be best if PCM were to charge completely in same amount of time it takes for

thermocline to pass. (The 2” diameter PCM tubes used as heat exchangers in the current

experimental set-up take longer to charge than it takes for the tank water to be replaced.) This goal

might be achieved through optimizing the material, size, shape, wall thickness, and location of the

heat exchangers in the tank. Simply changing the material of the heat exchanger could improve

the thermal conductivity of the tube walls and provide greater heat transfer. Future experiments

should be performed to examine whether smaller or large diameter tubes should be used. Though

other shapes of heat exchangers would be less affordable than tubes, the pros and cons of spherical

heat exchangers verses cylinders or the possibility of adding internal fins to the heat exchangers

should be investigated. The heat transfer between the PCM and the water might also be affected

by the heat exchanger’s location in tank. Heat exchangers located closer to the bottom of the tank

in charging would be exposed to the greatest temperature difference for the longest amount of time

possible, driving heat transfer. Likewise, heat exchangers located toward the top of the tank during

discharging would experience faster heat transfer. All of these factors will have to be taken into

account to produce a system that would be optimized for integration with the University of

Cincinnati’s full scale TES tanks.

73
Just as important to the optimization of internal heat transfer is the type of PCM used in the system.

When selecting a PCM to use in the University tanks, there must be thorough in house studies to

confirm the thermal conductivity, specific heat, latent heat, and chemical compatibility of the

PCM. Special attention should be paid to the freezing and melting points of the PCM because the

operating temperature range used in the University of Cincinnati water cooling system is relatively

small. It is desirable for the PCM to have a freezing and melting temperature that is halfway

between the average chilled water inlet temperature and the average return water temperature of

the system. It is also vital that the subcooling of the PCM be minimized as much as possible to

ensure consistent solidification in the intended time period. And last, chemical compatibility

between the PCM and the material of the heat exchanger should be considered. The PCM should

be a non-toxic material that will not cause excess corrosion of the heat exchanger and would not

harm the water quality in the case of a leak in the heat exchanger wall.

In support of these optimizations, thermal and mathematical modeling should be performed. This

should include the modeling of the heat transfer between the PCM inside the heat exchanger tube

and the water surrounding it. When there is a temperature difference between the water and the

PCM, it is possible for a thermal plume to form around the heat exchanger causing unwanted

mixing in the tank. It is important that these thermal plume effects also be modeled around tubes

during charging and discharging. Future modeling of the accurate internal heat transfer in the tank

and heat exchanger tubes will result in knowledge of how much mass of the PCM is in solid or

liquid phase at any given time in the charging and discharging process.

74
References

1. Wildin, M., Truman, C. (1985). Evaluation of Stratified Chilled Water Storage Techniques,

Vol. 1: Findings, Vol. 2: Appendices. Electric Power Research Institute Report EM-4352.

2. Wildin, M. (1990). Diffuser Design for Naturally Stratified Thermal Storage. ASHRAE

Transactions, 96(1), 1094-1102.

3. Truman, C., Wildin M. (1989). Finite Difference Model for Heat Transfer in a Stratified

Thermal Storage Tank with Throughflow. Numerical Heat Transfer with Personal

Computers and Supercomputing, ASME HTD, 110, 45-55.

4. Zurigat, Y., Maloney, K., Ghajar, A. (1989). A Comparison Study of One-Dimensional

Models for Stratified Thermal Storage Tanks. ASME Journal of Solar Energy Engineering,

111(3), 204-210.

5. Wildin, M. (1991). Flow Near the Inlet and Design Parameters for Stratified Chilled Water

Storage. ASME 91-HT, 27.

6. Bahnfleth, W. and Musser, A. (1998). Thermal Performance of a Full-Scale Stratified

Chilled-Water Storage Tank. ASHRAE Transactions, 104(2).

7. Stewart, W. (2001). Operating Characteristics of Five Stratified Chilled Water Thermal

Storage Tanks. ASHRAE Transactions, 107(2), 12-21.

8. Musser, A., Bahnfleth, W. (2001). Parametric Study of Charging Inlet Diffuser

Performance in Stratified Chilled Water Storage Tanks with Radial Diffusers: Part 1—

Model Development and Validation. HVAC&R Research, 7(2), 52-65.

9. Musser, A., Bahnfleth, W. (2001). Parametric Study of Charging Inlet Diffuser

Performance in Stratified Chilled Water Storage Tanks with Radial Diffusers: Part 2—

75
Dimensional Analysis, Parametric Simulations and Simplified Model Development.

HVAC&R Research, 7(2), 205-222.

10. Bahnfleth, W., Musser, A. (1999). Parametric Study of Charging Inlet Diffuser

Performance in Stratified Chilled Water Storage Tanks with Radial Diffusers. Final Report.

ASHRAE Research Project 1077.

11. Mackie, E. and Reeves G. (1988). Stratified chilled-water storage design guide, EPRI EM-

4852.

12. Mehling, H., Cabeza, L. (2008). Heat and Cold Storage with PCM. Berlin Heidelberg:

Springer-Verlag.

13. Zhang, Y., Jiang, Y. & Jiang, Y. (1999). A Simple Method, the T-History Method, of

Determining the Heat of Fusion, Specific Heat and Thermal Conductivity of Phase-Change

Materials. Measurement Science and Technology, 10, 201–205.

14. Dorgan, C., Elleson, J. (1993). Design Guide for Cool Thermal Storage. ASHRAE

Transactions.

15. Bergman, T., Levine, A., Incropera, F., Dewitt, D. (2011). Fundamentals of Heat and Mass

Transfer 7th edn. New York: Wiley.

16. Bejan, A. (1993). Heat Transfer. New York: Wiley).

17. Wildin, M., Mackie, E., Harrison W. (1990). Stratified Thermal Storage: A New/Old

Technology. ASHRAE Journal, (pp 29-40).

18. Grubbs J. (2012). University of Cincinnati West Campus Thermal Storage System

Operation Manual. Cincinnati: Fosdick & Hilmer.

76
19. Lazaro A., Gunther E., Mehling H., Hiebler S., Marin J., Zalba B. (2006). Verification of

a T-history installation to measure enthalpy versus temperature curves of phase change

materials. Measurement Science Technology, 17 2168-74.

20. Günther, E., Hiebler, S., Mehling, H. (2006). Determination of the Heat Storage Capacity

of PCM and PCM-Objects as a Function of Temperature. Proceedings of ECOSTOCK

Stockton USA.

21. Marin, J., Zalba B., Cabeza L., Mehling H. (2003). Determination of the Enthalpy-

Temperature Curves of Phase Change Materials with the T-History Method – Improvement

to Temperature Dependent Properties. Measurement Science Technology, 184 – 189.

22. Moffat, R. (1988). Describing the Uncertainties in Experimental Results. Experimental

Thermal and Fluid Science, 1, 3-17.

23. Yinping, Z., Yi, J., and Yi, J. (1999). A Simple Method, the T-History Method, of

Determining the Heat of Fusion, Specific Heat and Thermal Conductivity of Phase-Change

Materials. Measurement Science Technology, 10, 201-5.

77
Appendices

Appendix A: Additional Figures

Figure A.1: Original Diffuser Schematic and Photograph

78
Figure A.2: Thermoclines of Sensible Tests in 15 minute Increments Charging

240 180 min 165 120 min 105


min min min 75 min
105 min
150 min
90 min 60 min

120 min
75 min
45 min
90 min 60 min
30 min
45 min
60 min
15 min
30 min
30 min
0 min 15 min 0 min 0 min

Figure A.3: Thermoclines of Sensible Tests in 15 minute Increments Discharging

0 min 0 min
0 min

15 min
30 min
15 min

30 min
60 min
45 min 30 min

90 min 45 min
60 min

60 min
120 min 75 min
75 min
90 min
150 min 105 min

79
Figure A.4: Comparing Thermoclines of Nominally Augmented and Sensible Tests in 15
Minute Increments Charging

Nominal Augmented Test #22

90 min 90 min
75 min 75 min

60 min

60 min

45 min
45 min

30 min
30 min

15 min

15 min 0 min
0 min

Figure A.5: Comparing Thermoclines of Nominally augmented and Sensible Tests in 15


Minute Increments Discharging

0 min
0 min

15 min
15 min

30 min
30 min

45 min
45 min

60 min
90 min
60 min 75 min 75 min

80
Figure A.6: Testing Different Packing Factors of PCM

81
82
Appendix B: Matlab Codes

The “RUN” Code

Because calculations had to be run many times over for each different test data set, an overarching
code was made in order to call the function “TES Analysis” that would perform all the desired
calculations for the data analysis of each test. This “RUN” code is as follows:
%%%%%Run AAAALLLL the Data!!!!!! Create a spreadsheet with the results!
clear
close
clc

filename='TESCalc_25July2016.xlsx';

VFRMethod=1;
syms E dTwater
E=.1*dTwater;
%%Running data for sensible tests%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
sheet=1;
%%Sensible Charging

TESAnalysis_Sensible(xlsread('625C20140530.xlsx','A8:AD272'),0.8600,filename,
sheet,'B3',E,VFRMethod);

TESAnalysis_Sensible(xlsread('625sensible20140908to10.xlsx','A8:AD1299'),0.88
66,filename,sheet,'B4',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125C20140529.xlsx','A8:AD195'),1.4700,filename,
sheet,'B5',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125sensible20140916.xlsx','A8:AD171'),1.5739,fi
lename,sheet,'B6',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125sensible20140905.xlsx','A8:AD249'),1.8339,fi
lename,sheet,'B7',E,VFRMethod);

TESAnalysis_Sensible(xlsread('250sensible20140903.xlsx','A8:AD199'),2.1353,fi
lename,sheet,'B8',E,VFRMethod);

TESAnalysis_Sensible(xlsread('218C20140528.xlsx','A8:AD122'),2.2146,filename,
sheet,'B9',E,VFRMethod);

TESAnalysis_Sensible(xlsread('250sensible20140915.xlsx','A8:AD190'),2.2182,fi
lename,sheet,'B10',E,VFRMethod);

TESAnalysis_Sensible(xlsread('250C20140527_2.xlsx','A8:AD159'),2.2500,filenam
e,sheet,'B10',E,VFRMethod);

%%Sensibble Discharging

TESAnalysis_Sensible(xlsread('250sensible20140903.xlsx','A200:AD1228'),0.3906
,filename,sheet,'B13',E,VFRMethod);

83
TESAnalysis_Sensible(xlsread('625sensible20140908to10.xlsx','A1300:AD2550'),0
.9452,filename,sheet,'B14',E,VFRMethod);

TESAnalysis_Sensible(xlsread('625D20140530.xlsx','A8:AD335'),1.0400,filename,
sheet,'B15',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125sensible20140905.xlsx','A250:AD1454'),1.5854
,filename,sheet,'B16',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125sensible20140916.xlsx','A173:AD2201'),1.7267
,filename,sheet,'B17',E,VFRMethod);

TESAnalysis_Sensible(xlsread('125D20140529.xlsx','A8:AD175'),1.7300,filename,
sheet,'B18',E,VFRMethod);

TESAnalysis_Sensible(xlsread('250D20140527.xlsx','A8:AD103'),2.1500,filename,
sheet,'B19',E,VFRMethod);

TESAnalysis_Sensible(xlsread('250sensible20140915.xlsx','A193:AD593'),2.4545,
filename,sheet,'B20',E,VFRMethod);

TESAnalysis_Sensible(xlsread('218D20140528.xlsx','A8:AD185'),2.4600,filename,
sheet,'B20',E,VFRMethod);

%% Running data for nominally augmented


tests%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
sheet=2;
%%Nominally Augmented Charging

TESAnalysis_Nominal1(xlsread('DoubleChargeCopper625_20140626to28.xlsx','A8:AD
403'),0.7552,filename,sheet,'B3',E,VFRMethod);

TESAnalysis_Nominal2(xlsread('Single125C20140618.xlsx','A8:AD509'),1.2794,fil
ename,sheet,'B4',E,VFRMethod);

TESAnalysis_Nominal(xlsread('CompleteCopperCharge_20140609to10.xlsx','A8:AD13
78'),2.2146,filename,sheet,'B11',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitcharge_20140707.xlsx','A8:AD490'),2.
1971,filename,sheet,'B7',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitchargeWaitCharge_20140709.xlsx','A8:
AD772'),2.2189,filename,sheet,'B9',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitchargeWaitCharge_20140708.xlsx','A8:
AD750'),2.1808,filename,sheet,'B6',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitchargeWaitCharge_20140710to11.xlsx',
'A8:AD1719'),2.2200,filename,sheet,'B10',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitcharge_20140702.xlsx','A8:AD1443'),2
.1219,filename,sheet,'B5',E,VFRMethod);

%%Nominally Augmented Discharging

84
TESAnalysis_Nominal(xlsread('CompleteCopperDischarge250_20140623.xlsx','A8:AD
1386'),2.5104,filename,sheet,'B16',E,VFRMethod);

TESAnalysis_Nominal(xlsread('CopperDischarge125_20140625.xlsx','A8:AD1056'),1
.6524,filename,sheet,'B13',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitcharge_20140707.xlsx','A492:AD1380')
,2.4368,filename,sheet,'B14',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitchargeWaitCharge_20140709.xlsx','A77
4:AD1357'),2.4829,filename,sheet,'B15',E,VFRMethod);

TESAnalysis_Nominal(xlsread('2hchargeWaitchargeWaitCharge_20140708.xlsx','A76
0:AD1378'),2.6274,filename,sheet,'B17',E,VFRMethod);

The “TES Analysis” Code

A slightly modified version of this code was used to perform the necessary data analysis
calculations for each test. However the generalized code is as follows:
% This function calculates the Mass of Water and Mass of PCM present in the
% tank, the Mass Packing Factor, the Experimental and Theoretical
% Charge/Discharge Times, Temperature Difference, Average Thermocline
Thickness, Integrated and Maximum
% Capacity, and Half FOM and enters these values into output vectors.
function [vector,TTtable,Tin, TinitialBulk,data] =
TESAnalysis_Nominal(data,VFR,filename,sheet,xlRange,E, VFRMethod)

SetUp='9Cu';

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%% DEFINE CONSTANTS %%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

% ch2o : Specific heat of water


ch2o=4.2003; %kJ/kgC

% cPCM : Specific heat of water


cPCMlq=4.2003; %kJ/kgC
cPCMsolid=4.2; %kJ/kgC

% M : Mass of Water in Tank and Mass of PCM alone


rhoh2o=1000; %kg/m^3
rhoPCMlq=860;%kg/m^3
rhoPCMsolid=950;%kg/m^3
H=1.6256; %Water level in tank in meters (64 inches)
A=((29+3/16)/2)^2*pi*.00064516; %Cross sectional area of tank (m^2)
Vrack=.0045; % Volume of PVC rack submerged in m^3
Vheader=.00082685192; %Volume of two PVC headers in m^3

%Volume of the tubes depends on the configuration that was set-up.

85
switch SetUp
case 'Sensible'
Vtubes=0; %m^3
VPCMAlone=0;%m^3
case '9Cu'
Vtubes=.0135; %m^3
VPCMAlone=.01147;%m^3
case '4Cu'
Vtubes=.006; %m^3
VPCMAlone=.01147/9*4;%m^3
case '9Al'
Vtubes=.0198; %m^3
VPCMAlone=.01336;%m^3
case '2222'
Vtubes=.0138; %m^3
VPCMAlone=(.01147/9*2)+(.01336/9*2)+(.014275/16*2)+(.011943/9
*2);%m^3
case '16Al15'
Vtubes=.0208; %m^3
VPCMAlone=.014275;%m^3
case '9PVC'
Vtubes=.0171; %m^3
VPCMAlone=.011943;%m^3
case '4Al2+4Al15'
Vtubes=.014; %m^3
VPCMAlone=(.01336/9*4)+(.014275/16*4);%m^3
case '9Cu9Al'
Vtubes=.0333; %m^3
VPCMAlone=.01147+.01336;%m^3
otherwise
disp('you spelled your SetUp variable wrong');
end
V=H*A-Vrack-Vheader-Vtubes; %m^3
M=rhoh2o*V;%Kg water alone
MPCMAlone=VPCMAlone*rhoPCMlq;%kg

% Mass Packing Factor


PackingFactor=MPCMAlone/(MPCMAlone+M)*100;% in percent

% This section of code decides if data is charging or discharging and defines


appropriate constants accordingly.
if data(1,1)==1
mode='Charging'
inlet=25;%collum number to reference in data matrix
outlet=27;%collum number to reference in data matrix
Tc(1:2)=0;
for i=3:length(data)
Tin(i,1)=mean(data(2:i,inlet))%average chilled water temp
end
TinitialBulk=mean(data(1,12:20))%initial bulk temperature of tank

lastwaterRTD=data(1:length(data),12);
lastPCMRTD=data(1:length(data),3);
ChargeTimeWater=lastwaterRTD-Tin;
ChargeTimeWater(1:2)=1;
ChargeTimePCM=lastPCMRTD-Tin;

86
ChargeTimePCM(1:2)=1;
elseif data(1,2)==1
mode='Discharging'
inlet=26;%collum number to reference in data matrix
outlet=28;%collum number to reference in data matrix
Th(1:2)=0;
for i=3:length(data)
Tin(i,1)=mean(data(2:i,inlet));%average system return temp
end
TinitialBulk=mean(data(1,12:20));%initial bulk temperature of
tank

lastwaterRTD=data(1:length(data),20);
lastPCMRTD=data(1:length(data),11);
ChargeTimeWater=Tin-lastwaterRTD;
ChargeTimeWater(1:2)=1;
ChargeTimePCM=Tin-lastPCMRTD;
ChargeTimePCM(1:2)=1;
else
disp('not a charge or discharge. Check your data.')
end

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%% Calculatiions %%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%%%%%%Charge Time and Theoretical Charge TIme


ChargeTimeWater=min(find(ChargeTimeWater(1:end,1)<.5))-1%good
ChargeTimePCM=min(find(ChargeTimePCM(1:end,1)<.5))-1%good

% MFR : Mass flowrate through tank in m^3/min


if VFRMethod==2
VFR= levelmeterFR(data,ChargeTimeWater,0,4);
end
VFR2= VFR*.00378541178;%m^3/min
MFR=VFR2*rhoh2o; %kg/min

% RiseRate
Aft=((29+3/16)/2/12)^2*pi;%Area of tank in ft^2.
RiseRate=VFR*0.133681/Aft*12;%in/min

%Theoretical Charge Time


ChargeTimeTheory=round(M/MFR);%Time it takes to replace original mass
using average Mass flow rate.

%%%%%%Delta T for water and for PCM


TinletWater=Tin(ChargeTimeWater+1,1);
dTwater=abs(TinitialBulk-TinletWater);
E=subs(E,dTwater);
E=sym2poly(E);
E=E(1);
%dTPCM=abs(TiniialBulk-TinletPCM);

%%%%%%Average Thermocline Thickness


if data(1,1)==1

87
mode='Charging'
Tc=TinletWater
Th=TinitialBulk
elseif data(1,2)==1
mode='Discharging'
Tc=TinitialBulk
Th=TinletWater
else
disp('not a charge or discharge. Check your data.')
end

RTD=[outlet,13,14,15,16,17,18,19];
TTtable=zeros(1,10);

for n=1:8
[t1(1,n),t2(1,n),T1(1,n),T2(1,n),Tone(1,n),Ttwo(1,n),Temp,Time] =
...
FindingThermoclineV4_nograph(data,RTD(n),Tc,Th,mode,E);

TTref=[2,3,4,6,7,8];
TTtable(n)=abs(t2(n)-t1(n))*RiseRate;
TTtable(9)=mean(TTtable(TTref)); % average 80% thermocline
thickness
TTtable(10)=abs(TTtable(2)-TTtable(8));% how much does
thermocline thickness increase over course of process?

end
TTavg=TTtable(9)

%%%%%%Half FOM

dt=1;%minutes
Tinlet=F2C(data(1:ChargeTimeWater,inlet));
Toutlet=F2C(data(1:ChargeTimeWater,outlet));
CintWater=sum(ch2o.*MFR.*abs(Tinlet-Toutlet)*dt)./12660;%ton-hr

Tinlet2=F2C(data(1:ChargeTimePCM,inlet));
Toutlet2=F2C(data(1:ChargeTimePCM,outlet));
CintPCM=sum(ch2o.*MFR.*abs(Tinlet2-Toutlet2)*dt)./12660;%ton-hr

Cmax=MFR*(ChargeTimeWater)*ch2o*abs(F2C(Th)-F2C(Tc))/12660% ton-hr
CmaxPCM=MFR*(ChargeTimePCM)*ch2o*abs(F2C(Th)-F2C(Tc))/12660;% ton-hr
HalfFOM=CintWater/Cmax*100;
HalfFOMPCM=CintPCM/CmaxPCM*100;

if isempty(ChargeTimePCM)==1
ChargeTimePCM=0;
CmaxPCM=0;
HalfFOMPCM=0;
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%Ideal Sensible Integrated Capacity for same dT and flow rate
CmaxSensibleIdeal=MFR*(ChargeTimeTheory)*ch2o*abs(F2C(Th)-
F2C(Tc))/12660;% ton-hr

88
%%%%%%The Vector
vector=[VFR,dTwater,TTavg,ChargeTimePCM,ChargeTimeWater,ChargeTimeTheory,Half
FOMPCM,HalfFOM,CintPCM,CintWater,Cmax,CmaxSensibleIdeal];
vector(13:22)=fliplr(TTtable);

%%% excell sheet


A = vector;
xlswrite(filename,A,sheet,xlRange)
end

Supporting Functions

These functions are called on by the “TES Analysis” code above.

“FindingThermoclineV4”

function [t1,t2,T1,T2,Tone,Ttwo,Temp,Time] =
FindingThermoclineV4_nograph(data,RTD,Tc,Th,mode,E)
close all
%Delta T
dT=abs(Tc-Th);
RTD=RTD
%Defining %80 of the dT. The thermcline lies between these temperatures.
switch mode
case 'Charging'
T1=Th-(E);
T2=Tc+(E);
case 'Discharging'
T1=Tc+(E);
T2=Th-(E);
end
Temp=data(1:end,RTD);
Time=[1:1:length(Temp)];
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%Loop uses tlimit as initial guess for (time at which temperature is


%equal to Tc) and iterates to find tc
switch mode
case 'Discharging'
t1=find(Temp>T2,1)%t1=find(floor(Temp)==round(mean(Temp)),1)
if isempty(t1)==1
disp('alt t1')
t1=find(Temp>T2,1);%t1=find(ceil(Temp)==round(mean(Temp)),1)
end
Tone=T2;
while Tone-T1>=0 &&t1>=1&&t1<=length(Time)

Tone= interp1(Time,Temp,t1);
t1=t1;
switch mode
case 'Charging'
if Temp(t1)<T1&&T1<Temp(t1-1)
break

89
else t1=t1-1;
end
case 'Discharging'
if (Temp(t1)>T1&&T1>Temp(t1-1))
break
else t1=t1-1;
end
end
end
while Tone-T1>0 &&t1>=1&&t1<=length(Time)

Tone= interp1(Time,Temp,t1);
t1=t1;
switch mode
case 'Charging'
t1=t1-.0001;
case 'Discharging'
t1=t1-.0001;
end
if abs(Tone-T1)<=.0001
break
end
end

if t1<=1||t1>=length(Time)
disp('Use a different RTD for t1')
end
%Loop uses tlimit as initial guess for th (time at which temperature is
%equal to Th) and iterates to find th
t2=find(Temp>=T1,1);%t2=find(floor(Temp)==round(mean(Temp)),1)
if isempty(t2)==1
t2=1;
end

Ttwo=T1;
while T2-Ttwo>0 &&t2>=1&&t2<=length(Temp)
Ttwo= interp1(Time,Temp,t2)
t2=t2;
switch mode
case 'Charging'
if Temp(t2)>T2&&T2>Temp(t2+1)
break
else t2=t2+1;
end
case 'Discharging'
if Temp(t2)<T2&&T2<Temp(t2+1)
break
else t2=t2+1;
end
end
end

while T2-Ttwo>=0 &&t2>=1&&t2<=length(Temp)

Ttwo= interp1(Time,Temp,t2);
t2=t2;

90
switch mode
case 'Charging'
t2=t2+.0001;
case 'Discharging'
t2=t2+.0001;
end

if abs(Ttwo-T2)<=.0001
break
end

end
%disp('t2 loop ran')
if t2<=1||t2>=length(Time)
%disp('Use a different RTD for t2')
end

case 'Charging'

t1=find(Temp<T2,1);%t1=find(floor(Temp)==round(mean(Temp)),1)

if isempty(t1)==1
disp('alt t1')
t1=length(data);%t1=find(ceil(Temp)==round(mean(Temp)),1)

end
Tone=T2;
while Tone-T1>=0 &&t1>=1&&t1<=length(Time)

Tone= interp1(Time,Temp,t1);
t1=t1;
switch mode
case 'Charging'
if Temp(t1)<T1&&T1<Temp(t1-1)
%disp('first t1 loop ran')
break
else t1=t1-1;
end
case 'Discharging'
if (Temp(t1)>T1&&T1>Temp(t1-1))
%disp('first t1 loop ran')
break
else t1=t1-1;
end
end
end
while Tone-T1>0 &&t1>=1&&t1<=length(Time)

Tone= interp1(Time,Temp,t1);
t1=t1;

91
switch mode
case 'Charging'
t1=t1-.0001;
case 'Discharging'
t1=t1-.0001;
end
if abs(Tone-T1)<=.0001
break
end
end
%disp('t1 loop ran')
if t1<=1||t1>=length(Time)
%disp('Use a different RTD for t1')
end
%Loop uses tlimit as initial guess for th (time at which temperature is
%equal to Th) and iterates to find th
t2=find(Temp<T1,1);%t2=find(floor(Temp)==round(mean(Temp)),1)
if isempty(t2)==1
%disp('alt t2')
t2=find(Temp>=T1,1);%t2=find(ceil(Temp)==round(mean(Temp)),1)
end

Ttwo=T1;
while T2-Ttwo>0 &&t2>=1&&t2<=length(Temp)
Ttwo= interp1(Time,Temp,t2);
t2=t2;
switch mode
case 'Charging'
if Temp(t2)>T2&&T2>Temp(t2+1)
%disp('first t2 loop ran')
break
else t2=t2+1;
end
case 'Discharging'
if Temp(t2)<T2&&T2<Temp(t2+1)
%disp('first t2 loop ran')
break
else t2=t2+1;
end
end
end
%disp('end first loop')
while T2-Ttwo>=0 &&t2>=1&&t2<=length(Temp)

Ttwo= interp1(Time,Temp,t2);
t2=t2;
switch mode
case 'Charging'
t2=t2+.0001;
case 'Discharging'
t2=t2+.0001;
end
if abs(Ttwo-T2)<=.0001
break
end
end

92
%disp('t2 loop ran')
if t2<=1||t2>=length(Time)
%disp('Use a different RTD for t2')
end

end

end

“levelmeterFR”

%This code uses the time record of flowrate from the PLC controler,
eliminates bad data points, and calculates the average flow rate
function [ flowrate,data ] = levelmeterFR(
data,ChargeTimeWater,lowlim,highlim )
%UNTITLED8 Summary of this function goes here
% Detailed explanation goes here
highlim=3;
data=data(1:ChargeTimeWater+1,24);
figure (1)
hist(data,[min(data):.1:max(data)]);
pause(5);
r=find(data(1:length(data),1)>lowlim);
data=data(r,1);
figure (2)
hist(data,[min(data):.1:max(data)]);
pause(5);
r=find(data<highlim);
data=data(r);
figure (3)
hist(data,[min(data):.1:max(data)]);
pause(5);
flowrate=mean(data);
end

93
Appendix C: Properties of PureTemp 8

94
Table A.1: Cost of PCM Only
Gallons of PCM Required at Price Per Gallon Cost of PCM

1.6% Mass Packing Factor

LabTES Tank 3 Gallons $10.91 $32.73

Power Plant Tank 55,669 Gallons $10.91 $607,348.79


(3 million Gallons)

95
Appendix D: Addition Thermal Capacity Calculations
The following Matlab code was used to calculate the thermal capacity of the PVC rack, the PVC

distribution rings (headers), and the copper heat exchangers in the augmented tank.

Code
clear
clc
%Calculating the mass of the PVC rack and PVC headers...
Vrack=.0045; % Volume of PVC rack submerged in m^3
Vheader=.00082685192; %Volume of two PVC headers in m^3
rhoPVCrack=1450;%kg/m^3 Rigid PVC
rhoPVCheaders=1350;%kg/m^3 Flexible PVC
MassRack=rhoPVCrack*Vrack;%kg
Massheaders=rhoPVCheaders*Vheader;%kg

%Calculating the mass of the copper tubes. Vtubes was found using a water
%displacement test...

Vtubes=.0135; %m^3
VPCMAlone=.01147;%m^3
Vcopper=Vtubes-VPCMAlone;%m^3
rhoCu=8940;%kg/m^3
MassCu=rhoCu*Vcopper;%kg

%Calculating the thermal storage capacity of the copper tubes, the


%headers, and the rack...

%Specific heat values for copper and PVC


Ccopper=.39;%kJ/kgC
CPVC=1;%kJ/kgC
%Defining the average temperature difference
dT=6;% degrees Celsius

%Calculating the thermal storage capacity in Ton-hours


StorageCapacityRack=MassRack*CPVC*dT/12660;%Ton-Hours
StorageCapacityDistributionRings=Massheaders*CPVC*dT/12660;%Ton-Hours
StorageCapacityTubes=MassCu*Ccopper*dT/12660;%Ton-Hours
sum=StorageCapacityRack+StorageCapacityHeaders+StorageCapacityTubes;%Ton-
Hours

Results

StorageCapacityRack = 0.0031 Ton-Hours

StorageCapacityDistributionRings = 5.2903e-04 Ton-Hours

96
StorageCapacityTubes = 0.0034 Ton-Hours

sum = 0.007 Ton-Hours = 0.4509% of the theoretical thermal capacity of the augmented tank

(1.547 Ton-Hours as mentioned in Chapter 3)

The PVC rack with distribution rings and the copper tubes both contribute about equally to the

thermal capacity of the tank. However, both contributions are very small and negligible in

comparison to the capacity of the water and PCM in the tank.

97

You might also like