Scholarly article on topic 'Experimental and Numerical Investigation of Stability of Packed Bed Thermal Energy Storage for CSP Power Plant'

Experimental and Numerical Investigation of Stability of Packed Bed Thermal Energy Storage for CSP Power Plant Academic research paper on "Materials engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Energy Procedia
OECD Field of science
Keywords
{"thermal energy storage" / thermocline / "rock bed" / "sensible heat" / "CSP power plant" / "numerical simulation"}

Abstract of research paper on Materials engineering, author of scientific article — A. Bruch, J.F. Fourmigue, R. Couturier, S. Molina

Abstract This paper is focused on experimental and numerical investigation of a dual-media rock bed thermal energy storage. An experimental test loop whose main component is a 1 m diameter and 3 m high storage tank has been built to carefully study the thermal behavior of rock bed thermocline thermal storage. Experimental results show constant outlet temperature during discharge process and confirm that thermocline is suitable for integration in a CSP power plant. The influences of mass flow rate, charge and discharge temperature difference and partial load have been tested without significant effect on dimensionless axial temperature profile. This indicates the robust and controllable nature of dual-media thermocline energy storage. Numerical 3D simulations have been performed using the commercial code Fluent. Results clearly illustrate that the piston-like behavior of thermocline can be strongly degraded by flow non-uniformities. First comparisons with experimental data show global good agreement even if some improvements of the model are needed.

Academic research paper on topic "Experimental and Numerical Investigation of Stability of Packed Bed Thermal Energy Storage for CSP Power Plant"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Energy Procedía 49 (2014) 743 - 751

SolarPACES 2013

Experimental and numerical investigation of stability of packed bed thermal energy storage for CSP power plant

A. Brucha, J.F. Fourmiguea, R. Couturiera and S. Molinaa

a CEA/INES, Laboratoire des Systèmes Thermiques (LETh), BP 332, 50 avenue du Lac Léman, F-733 77 Le-Bourget-du-lac

Abstract

This paper is focused on experimental and numerical investigation of a dual-media rock bed thermal energy storage. An experimental test loop whose main component is a 1 m diameter and 3 m high storage tank has been built to carefully study the thermal behavior of rock bed thermocline thermal storage. Experimental results show constant outlet temperature during discharge process and confirm that thermocline is suitable for integration in a CSP power plant. The influences of mass flow rate, charge and discharge temperature difference and partial load have been tested without significant effect on dimensionless axial temperature profile. This indicates the robust and controllable nature of dual-media thermocline energy storage. Numerical 3D simulations have been performed using the commercial code Fluent. Results clearly illustrate that the piston-like behavior of thermocline can be strongly degraded by flow non-uniformities. First comparisons with experimental data show global good agreement even if some improvements of the model are needed.

© 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/3.0/).

SelectionandpeerreviewbythescientificconferencecommitteeofSolarPACES2013underresponsibilityofPSEAG. Final manuscript published as received without editorial corrections.

Keywords: thermal energy storage; thermocline; rock bed; sensible heat; CSP power plant; numerical simulation

1. Introduction

Thermal energy storage (TES) systems offer the potential to increase CSP power plant effectiveness by extending or delaying the electricity production period and by correcting the mismatch between the supply and the demand of energy ([ 1 ]). Up to now, two tanks TES system based on the use of a cold tank and a hot tank is the most reliable technology which has already been deployed in large scale in CSP power plant. EPRI ([ 2 ]) recently conducted a comparative study between two tanks TES and single tank thermocline TES and concluded that dual-media thermocline TES using a mixture of rock as primary storage material would have a potential to reduce about 33 % the cost of thermal energy storage.

1876-6102 © 2013 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/Kcenses/by-nc-nd/3.0/).

Selection and peer review by the scientific conference committee of SolarPACES 2013 under responsibility of PSE AG. Final manuscript published as received without editorial corrections. doi: 10.1016/j.egypro.2014.03.080

Only few studies are devoted to experimental investigation of rock bed thermocline using liquid heat transfer fluid. The concept was demonstrated and validated at industrial scale with a 170 MWhth tank on the Solar One pilot central receiver ([ 3 ]) with rocks and thermal oil. More recently, Sandia studied a smaller-scale 2.3 MWhth thermocline using molten salt as heat transfer fluid ([ 4 ]). Both studies exhibit typical thermocline temperature profiles but significant experimental scattering was observed. Mawire et al. ([ 5 ], [ 6 ]) experimentally investigated the thermal response of a single-tank rock/oil thermocline at a lab-scale using a small tank of 30 cm high. Numerous publications present numerical simulation of rock bed thermocline and try to evaluate the behavior of such storage system under different operating conditions ([ 7 ], [ 8 ]).

Despite these studies reported in the literature, the thermal behavior and the effectiveness of dual-media thermocline are not well understood. In particular, the stability of the thermocline behavior under various operating conditions is not well documented at this time. Thus a dedicated experimental test loop at a significant scale is needed to exhaustively investigate the thermocline characteristics in large ranges of operating parameters for later design for implantation in CSP power plants. In parallel a 3D CFD numerical simulation has been developed.

Nomenclature

Cp Specific heat in J/kg.K

D Diameter in m

E Energy in kWh

E* Dimensionless energy

L Length in m

m Mass flow rate in kg/s

T Temperature in °C

T* Dimensionless temperature

t Time in s

* t Dimensionless time

U Velocity in m/s

V Volume in m3

z Axial position in m

z Dimensionless axial position

P Density in kg/m3

charge refers to charge

discharge refers to discharge

f refers to fluid

sup superficial

tank refers to tank properties

2. Experimental set-up

To carefully investigate and evaluate the dual-media thermocline behavior and performances at a representative scale, a specific experimental installation has been built and exhaustively instrumented. An overview of the flow loop is presented in Fig. 1. The experimental set-up consists of a storage tank and two fluid loops to simulate the charging and discharging components, i.e. the solar field and the electricity production device, respectively. The thermocline tank is a cylindrical stainless steel tank of 1 m diameter and 3 m high surrounded by a dome top and a dished bottom used as fluid distributors. No specific fluid distributor are added as the dished bottom and the dome top are supposed to be sufficient to provide uniform flow distribution at the pebble bed inlet. The rock bed is composed of silica rocks and silica sand mixture to minimize the void fraction. Non-pressurized thermal oil is used

as heat transfer fluid. The charging loop is composed of an electric heater, a regulated pump, regulation valves and a mass flow meter. The discharging loop integrates a regulated pump, regulation valves, mass flow meters and air cooler. The experimental setup was designed to allow dissociating the nominal conditions during charge and discharge, i.e. temperatures taken at the top of the storage and at the inlet of the air cooler can be different. This situation corresponds to the specific case of a CSP power plant where the storage temperature is higher than the maximum operating temperature of the turbine.

VR100/200-1

VR100/200-2

Fig. 1. Experimental test loop.

The temperature distribution in the storage tank was carefully investigated using about 250 K-type thermocouples (TC) placed both in the rocks and in the thermal oil. The different TC positions allow to accurately follow the thermocline through the tank and to investigate the perturbations in the axial and transverse temperature distributions. With the calibration and the automation, the uncertainty on the flow rates and the temperature measurements were less than 1% and ± 2°C.

3. Experimental results

3.1. Thermal behavior during discharge process

Fig. 2 shows the time-evolution of the dimensionless axial temperature profile in the rock bed during a discharge process from a fully charged tank. The results are presented using dimensionless variables defined as:

(Tf Ttdischarge)

{^charge ^discharge)

t* = —,-r- with = ■

sup nD2

, D tank

Lfank /1 Usup

z* =-- and E* =

PfCpfVtank(Tcharge Tdischarge

where Tf is the fluid temperature, Tcharge and Tdischarge are the charge and discharge temperatures defined by the operator, t is the time, z is the axial position in the storage tank, m is the mass flow rate, Dtank is the tank diameter, Ltank is the tank height and E is the energy. On Fig. 2 typical thermocline profiles, characterized by a hot region and a cold region separated by a moving thermal gradient, are observed. As reported experimentally by [ 4 ] and numerically by [ 9 ] and [ 10 ], the thermocline region is expanding with the discharging time. Results presented in Fig. 2 confirm the validity and the reliability of the instrumentation installed in the storage tank to follow the thermocline front during charge/discharge process.

As shown in Fig. 3, the oil outlet tank temperature is almost constant during most of the discharge. The measured temperature drop is less than 10°C at t =1.45 and less than 15°C at t =1.7. From this moment, the temperature gradient, or thermocline, is moving out of the tank and the outlet tank temperature is decreasing sharply and quickly: the temperature drop reaches 55°C at t =1.93. It should be noted that a slight time lag exists between the rock bed temperature and the outlet tank temperature as can be seen at a given time from Fig. 2 and Fig. 3. This is due to the dome top fluid and it does not impact significantly the experimental results. In Fig. 3, the dimensionless energy E is calculated as the ratio of the extracted energy and the maximum energy that can be stored in the tank at the used temperatures levels. Due to almost constant mass flux, inlet and outlet temperatures, the dimensionless energy E extracted from the tank evolves almost linearly during the whole discharge. A slope reduction is observed at the end of the discharge as the result of the diminution of the outlet temperature. The useful energy (i.e. from nominal turbine temperature to minimum temperature turndown of the turbine) extracted from the tank was approximately E*=0.89.

0.1 -0

t*=0.24

t*=0.48

t*=0.73

t*=0.97

■ A - t*=1.2

t*=1.45

t*=1.7

t*=1.93

Fig. 2. Dimensionless fluid temperatures in the storage tank during the discharge process.

These results confirm the ability of our dual-media thermocline tank to match the requirements of a CSP power plant storage system. A constant discharging temperature and discharged heat duty are of main interest in CSP plants to ensure proper functioning and more particularly when a turbine is used for electricity production, as the efficiency of such system notably decreases with temperature or load variation.

Fig. 3 : Dimensionless inlet and outlet tank temperatures and dimensionless extracted energy during the discharge process.

3.2. Influence of mass flow rate on thermocline temperature profiles

Influence of the mass flow rate on the thermocline has been experimentally tested. From a given initial condition of full storage tank, a specific charge/discharge cycle has been performed for four different mass flow rates. The end of the charge (respectively discharge) is defined when the mean transversal temperature reach a given value defined by the operator at the top (respectively at the bottom) of the rock bed. The inlet Reynolds numbers, based on Dtank and Usup, were varying from 9 to 24 in charge and from 1.5 to 3.6 in discharge. Fig. 4 presents the temperature profiles at the end of the charge/discharge process. In discharge mode, taking into account the temperature measurement uncertainty, a single behavior is observed indicating that the mass flow rate does not significantly impact the development of the thermocline. One can conclude the same in charge mode as the differences shown in Fig. 4 at the lowest mass flux are only observed in the high temperature region of the storage tank and can be explained by higher thermal losses due to higher residence time.

Fig. 4 : Influence of the inlet Reynolds number on the temperature profiles at the end of the charge/discharge process.

3.3. Influence of temperature level on thermocline temperature profiles

The influence of the charge and discharge temperature difference on the thermocline development has been tested for values of (Tchalge-Tdischarge) of 100°C and 150°C. The end-cycle criterions have been adapted to ensure same dimensionless temperature at the end of the charge and of the discharge. Fig. 5 shows that dimensionless axial thermocline temperature profiles are independent from the charge and discharge temperatures and that single dimensionless representation can be used. This indicates that experimental tests performed in given conditions may be extrapolated with any degree of certainty.

Fig. 5 : Influence of temperature levels on thermocline axial temperature profiles.

3.4. Influence of partial load on thermocline temperature profiles

The influence of a significant perturbation on an established charge/discharge cycle has been experimentally investigated through partial load. From an initial condition of full storage, rock bed is submitted to repetitive charge/discharge cycles. When established thermocline behavior is reached, a partial load is applied whereas discharge condition is kept constant.

Fig. 6 : End-cycle temperature profiles before, during and after the partial load cycle.

Fig. 6 presents the end-cycle rock bed temperature profiles for the partial load cycle and for the cycles before and after it. No difference are observed on the end-cycle temperature profile before and after the partial load, both in charge and in discharge, indicating that such perturbation does not influence significantly the thermocline behavior. It has to be added that the partial load has no significant impact on the discharge following it when constant discharge condition is applied. This indicates that dual-media thermocline energy storage has a robust behavior which is not degraded by intermittent perturbation.

4. Numerical results

Three dimensional simulation of the rock bed thermal energy storage has been performed using the commercial CFD software Fluent. The numerical geometry exactly fits the real tank, including the specific designs of the wall, the dome top, the dished bottom, the fluid inlets and outlets etc... Properties of the thermal oil, rock and sand are considered to be temperature-dependent. The rock bed is treated as an isotropic porous medium whose permeability was based on pressure drop measurements made on the experimental set-up.

4.1. Flow distribution

Fig. 7 illustrates the thermal front propagation for a quite correct and a degraded thermocline during a single charge. In Fig. 7 (a), the calculation conditions, in terms of fluid velocity, fluid temperature and rock bed pressure drop, correspond to the experimental values of one test presented on Fig. 4. A piston-like propagation of the thermal front is observed with only moderate transversal temperature gradients. During the charge, the thermal front is slightly destabilized by thermal losses to the ambient and flow non-uniformity in the rock bed due to jet effects in the inlet plenum. Fig. 7 (b) illustrates the fluid flow inside the storage tank in the case of a hypothetical situation of high fluid velocity and low rock bed pressure drop. Piston-like thermal front is no more observed and fluid flow is characterized by important preferential paths and large transversal temperature gradients. In this situation, the thermal energy storage system can no longer be called thermocline and the tank outlet temperature is no longer controlled. Such situation can be avoided by careful filling of the packed bed, specific care of the hydraulic distribution at the inlets of the storage tank and limitation of the maximum fluid velocity in the rock bed. These points have to be taking into account, when designing a dual-media thermocline storage, for proper integration in an industrial process such as CSP power plant.

IiiSSS

Fig. 7 : Propagation of the thermal front. (a): correct thermocline; (b): degraded thermocline.

4.2. Comparison with experimental data.

Fig. 8 presents numerical axial profiles of mean transversal temperature compared with experimental data for the charge test of Fig. 7. A global good agreement is obtained between CFD simulation and measurements. The numerical thermocline seems to run faster than in the experiments which may be corrected by improvement of thermal losses calculations, rock bed properties evaluation and/or internal wall conditions. In addition, numerical results exhibit lower axial thermal gradient, indicating that the heat transfer between the fluid and the rock bed is under-estimated. Generally speaking, the good agreement at t =0.2 indicates that the dome top and the dished bottom are correctly simulated.

0 0.2 0.4 0.6 0.8 i

Fig. 8 : Comparison of numerical and experimental temperature profiles during charge process.

5. Conclusion

A specific experimental installation has been built to study the hydraulic and thermal behaviors of dual-media thermocline using thermal oil as heat transfer fluid and a mixture of rock and sand as primary storage material. The storage tank is 1 m diameter and 3 m high and is instrumented with more than 250 thermocouples to finely follow the thermal front. No experimental scattering is observed and typical thermocline profiles characterized by a hot region and a cold region separated by a thermal gradient are obtained. Results show that thermocline energy storage behaves as expected as constant heat duty and outlet temperature are observed in discharge mode. The influences of mass flow rate, charge and discharge temperature difference and partial load have been tested without significant effect on axial temperature profile. In addition, all these tests show that thermocline behavior is highly reproducible. This indicates the robust and controllable nature of dual-media thermocline energy storage and confirms that such storage is suitable for implementation in a CSP power plant.

A 3D numerical model of the previously mentioned storage tank has been developed using the commercial code Fluent. Results confirm that, in the ranges of operating parameters of the present study, no significant bypass or preferential paths exist in the storage tank. On the contrary, results performed with highly out of range parameters show that thermocline behavior can be strongly degraded. Comparisons made with experimental data show global good agreement.

References

[ 1 ] Gil A., Medrano M, Martorell I et al, State of the art on high temperature thermal energy storage for power generation Part 1-Concepts, materials and modellization, Renewable Sustainable Energy Rev 2010, 14:31-55.

[ 2 ] Electric Power Research Institute. Solar Thermocline Storage Systems: Preliminary Design Study. Palo Alto, CA ( 2010) 1019581.

[ 3 ] Faas, S.E., Thorne, L.R., Fuchs, E.A., Gilbertsen, N.D., 1986. 10 MWe Solar Thermal Central Receiver Pilot Plant: Thermal Storage System Evaluation - Final Report. Sandia National Laboratories, Report, SAND86-8212.

[ 4 ] Pacheco JE, Showalter SK, Kolb WJ, Development of a molten salt thermocline thermal storage system for parabolic through plants, Journal of Solar Energy Engineering, 124, 153-159.

[ 5 ] Mawire A., Mc Pherson M., Experimental and simulated temperature distribution of an oil-pebble bed thermal energy storage system with a variable heat source, Applied Thermal Engineering 29 (2009) 1086-1095.

[ 6 ] Mawire A., Taole S., A comparison of experimental thermal stratification parameters for an oil-pebble-bed thermal energy storage (TES) system during charging, Applied Energy, 88 (2011) 4766-4778.

[ 7 ] Kolb W.J., 2006. Performance analysis of thermocline energy storage. In: ISEC2006-99005. Proceedings of the ASME International Solar Energy Conference, Denver, Colorado.

[ 8 ] Flueckiger SM, Garimella SV, Second-lax analysis of molten-salt thermal energy storage in thermocline, Solar Energy, In Press, 2012.

[ 9 ] Van Lew J.T., Chan C.L., Karaki W., Stephens J., Analysis of heat storage and delivery thermocline tank having solid filler material, Journal of solar energy engineering, 133 (2011) 021003-1-021003-10.

[ 10 ] Xu C., Wang Z., He Y., Li X., Bai F., Sensitivity analysis of the numerical study on the thermal performance of a packed-bed molten salt thermocline thermal storage system, Applies Energy 92 (2011) 65-75.