Concentrating solar power (CSP) is a renewable

Solar Selective Coatings for Concentrating Aaron Hall* Andrea Ambrosini Clifford Ho Sandia National Laboratories Albuquerque, N. Mex. Thermal spray c...
Author: Prudence Rose
7 downloads 3 Views 175KB Size
Solar Selective Coatings for Concentrating Aaron Hall* Andrea Ambrosini Clifford Ho Sandia National Laboratories Albuquerque, N. Mex.

Thermal spray coatings and solution-based synthesized coatings show promise as more efficient, durable materials than those currently used in concentrating solar power receiver applications.

*Member of ASM International and member, ASM Thermal Spray Society

C

oncentrating solar power (CSP) is a renewable energy technology that converts solar thermal energy to mechanical work via a heat engine, which is then converted to electricity through a generator. These systems are typically large—capable of generating tens to hundreds of megawatts of electricity. Nearly 500 MW of concentrating solar power are currently installed in the U.S. CSP systems use numerous mirrors to reflect and concentrate the sunlight onto receivers that heat a working fluid. Several mirror configurations are possible, including dishes, parabolic troughs, linear Fresnel, and heliostats. One of the most promising CSP technologies is the central receiver (or power tower) system, which consists of a field of large, nearly-flat mirror assemblies (heliostats) that track the sun and focus the sunlight onto a receiver on top of a tower (Fig. 1). In a typical configuration, a heattransfer fluid such as water/steam or molten salt is heated in the receiver and used to power a conventional steam-turbine Rankine cycle to generate electricity. Excess thermal energy collected in molten salts can be stored in large insulated tanks allowing operation of the steam turbine during the night or on cloudy days. The efficiency of a power tower can be increased if the energy absorbed by the receiver is maximized while the heat loss from the receiver to the environment is minimized. When a material heats up, energy is radiated in the infrared wavelengths. This phenomenon is known as thermal emittance and represents a heat loss for the CSP system. Thus, heat loss occurs because of thermal emittance from the hot receiver surface to the environment, as well as convection due to wind and buoyancy effects.

Higher central receiver operating temperatures (>600°C) are needed to improve power cycle efficiency and lower the cost of solar generated electricity. However, higher operating temperatures result in increased energy loss due to thermal emittance. Therefore, improved selective absorber coatings are an important part of CSP receiver development. An ideal selective absorber coating for CSP receivers would have high absorptivity in the solar spectrum to maximize energy capture at the receiver and a low emissivity in the infrared spectrum to minimize thermal radiative losses. For CSP systems to meet an electricity cost target of $0.06/kWh [1], new materials capable of extended operation at temperatures above 600°C are needed. Ideally, these materials will have high absorptance (> 0.95) in the solar spectrum (~250-2500 nm) and low thermal emittance (< 0.05) in the infrared spectrum (~1.5-20 μm at an emittance temperature of ~600°C). Note that there is some overlap in these solar and thermal spectra, which makes the development of selective properties challenging. In addition, the materials need to be stable in air, low-cost, easily applied at large scales in the field, and capable of surviving thousands of heating and cooling cycles. Currently, Pyromark Series 2500 high-temperature paint is the standard for CSP central receivers. It has a measured solar absorptance of 0.96, is low cost, and is easily applied. However, with a thermal emittance of 0.86, it suffers from large thermal losses during high temperature operation. It also degrades over time when operated in air causing a decline in performance and added operating costs for CSP facilities. Research at Sandia National Laboratories

Fig. 1 — A field of heliostats (mirrors) surrounds the Concentrating Solar Power Tower Central Receiver at the National Solar Thermal Test Facility at Sandia National Laboratories in Albuquerque, N. Mex. Courtesy of Randy Montoya (SNL). 28

ADVANCED MATERIALS & PROCESSES • JANUARY 2012

Solar Power Central Receivers

Coating characterization Solar absorptance (α) was measured using a solar spectrum reflectometer weighted to provide a measurement spectrum that closely approximates the air mass solar spectrum. A white diffuse standard (α = 0.198) was used for calibration. Thermal emittance (ε80°C) measurements were performed using an infrared reflectometer with an 80°C black body source. A gold standard (ε = 0.02) and a black standard (ε = 0.908) were used to calibrate the instrument. Due to instrument limitations, values given below for emissivity are assumed to have a ±10% error. Diffuse reflectance (absorbance) was taken at room temperature using a spectrophotometer from wavelengths of 200-2400 nm. A BaSO4 reference standard was used for calibration. Test coupon performance was ranked using a figure of merit (FOM) defined as: FOM (W/cm2) = 60αsolar – 5[(ε80°C + ε2400nm)/2] where αsolar, ε80°C, and ε2400nm are the solar absorptance, emittance at 80°C, and emittance at 2400 nm, respectively. The constants 60 and 5 have the units (W/cm2) and represent the energy flux incident on a central receiver and the

45 40 35 30 25 20 15 10

WC-9Co_P

WC-9Co

WC-20Co_P

WC-20Co

Tungsten_P

Tungsten

Ni-5Al_P

Ni-5Al

0

Ni-25Graphite_P

5 Ni-25Graphite

Coating preparation Thermal spray technology offers the ability to rapidly prepare thick (>1 mm) ceramic and metal coatings in the field. Sandia applied thermal spray coatings on 304L stainless steel using an air plasma spray (APS) torch using a number of commercially available thermal spray feed stock materials. Detailed spray process conditions can be found in Ref. 2. Solution-based approaches (spin coating and dip coating) were used to prepare spinel coatings. These techniques allow for considerably more flexibility in coating composition than thermal spray techniques. Dopants can be incorporated in spinel films by adding species to the aqueous precursor solutions. Both spin and dip coating techniques involve preparation of aqueous precursor solutions containing metal nitrates and a wetting agent (Triton X). Solution precursors for spin coating also use citric acid as a complexing agent. A thin layer of solution is applied to a 304L substrate using spin or dip coating, and the coated substrate is dried and sintered at high temperature (500 or 600°C) for up to six hours to burn off nitrates and organics, forming the spinel phase. The process can be repeated multiple times to build coatings of the desired thickness. Specific details of the solution based coating preparation can be found in Ref. 3.

50

Figure of merit, W/cm2

that addresses the issue of more efficient, durable solar selective materials for CSP receiver applications with coatings prepared using thermal spray and solution-based synthesis techniques is discussed in this article.

Fig. 2 — Figure of merit (FOM) values for as-deposited (filled bars) coating test coupons and polished coating test coupons with a 1 μm surface finish (open bars).

energy flux emitted by a blackbody at 700°C, respectively. The emittance term provides an estimate of the average emittance over the wavelength spectrum of interest. The emittance at 2400 nm was calculated from the diffuse reflectance data by assuming zero transmission through the sample and by applying Kirchoff ’s law. The FOM reflects the idea that maximizing absorptance at the central receiver does more to improve receiver efficiency than minimizing thermal emittance. The magnitude of energy absorbed by the receiver depends directly on the energy flux magnitude incident upon the surface; whereas, the magnitude of energy emitted by the receiver is only affected by the incident radiation if this flux leads to an increase in the receiver body temperature. Additionally, receiver materials are opaque to solar energy; therefore, maximizing the receiver absorptance minimizes the reflectance from the receiver surface. Measured thermal radiative properties for Pyromark Series 2500 are αsolar = 0.964, ε80°C = 0.862, and ε2400nm = 0.960, which equates to a FOM of 53.3. Each data set was obtained by making measurements on multiple samples taken from the same coating. For samples containing data with an uncertainty interval, the number of samples within the data set ranged from two to five. Data without an uncertainty interval indicate that only one sample was tested for that condition. Uncertainty intervals (Δ), where shown, were calculated according to: Δ = t0.75,n-1(σ/n1/2) where σ is the standard deviation of the data set, n is the ADVANCED MATERIALS & PROCESSES • JANUARY 2012

29

TABLE 1 — ABSORPTANCE AND EMITTANCE DATA FOR 304L STAINLESS STEEL TEST COUPONS Test coupon

Surface finish

Solar absorptance

Emittance @80°C

Emittance @2400 nm

Figure of merit, W/cm2

A

As-purchased

0.47

0.24

NA

NA

C(a)

As-purchased

0.87

0.35

NA

NA

D

Grit blasted

0.67

0.46

0.52

37.7

E(b)

Grit blasted

0.83

0.71

0.60

46.6

F

1 μm polish

0.15

0.10

NA

NA

G

1 μm polish

0.17

0.11

NA

NA

(a) Heat-treated for 12 h at 800°C before testing. (b) Heat-treated for 6 h at 600°C before testing.

TABLE 2 — ABSORPTANCE AND EMITTANCE DATA FOR COATING TEST COUPONS WITH AS-SPRAYED SURFACES Solar absorptance

Emittance @80°C

Emittance @2400 nm

Figure of merit, W/cm2

Ni-25 graphite

0.81±0.00

0.62±0.03

0.66±0.05

45±0

Ni-5Al

0.63±0.02

0.39±0.06

0.47±0.02

35±2

WC-20Co

0.82±0.03

0.55±0.03

0.61±0.09

46±4

WC-9Co

0.73±0.05

0.53±0.03

0.56±0.02

40±7

WC-25Ni

0.75±0.04

0.51±0.03

0.56

39

Co-28Mo17.5Cr-3.5Si

0.79±0.00

0.47±0.00

0.62±0.03

44±0

Coating

TABLE 3 — ABSORPTANCE AND EMITTANCE DATA FOR COATING TEST COUPONS WITH 1 μM POLISHED SURFACES Coating

Solar absorptance

Emittance @80°C

Emittance @2400 nm

Figure of merit, W/cm2

Ni-25 graphite

0.52±0.07

0.33±0.06

0.82

26

Ni-5Al

0.26±0.06

0.11±0.05

0.92

12

WC-20Co

0.60

0.32

0.82

37

WC-9Co

0.45±0.16

0.24±0.07

0.88

28

WC-25Ni

0.60±0.09

0.24±0.03

0.88

35

TABLE 4 — ABSORPTANCE AND EMITTANCE DATA FOR COATING TEST COUPONS WITH AS-SPRAYED SURFACES HEAT TREATED FOR 6 H AT 600°C Coating Ni-25 graphite

30

Solar absorptance

Emittance @80°C

Emittance @2400 nm

Figure of merit, W/cm2

0.93

0.78

0.81

52

Ni-5Al

0.89

0.57

0.75

50

Co-28Mo17.5Cr-3.5Si

0.86

0.57

0.70

48

ADVANCED MATERIALS & PROCESSES • JANUARY 2012

number of values in the data set, and t0.75,n-1 is the critical value for capturing 75% of a two-sided t-distribution used to describe the data set. Properties of thermal spray coatings Measured optical property data for each thermal spray coating are presented in Tables 1-4. The effects of surface roughness and heat treatment were also evaluated. Effect of surface roughness: Data in Tables 1-3 indicate that reducing the surface roughness lowered both the solar absorptance and emittance values. Such decreases are consistent with an increase in reflectance of light waves reaching the surface from both outside and within the samples. Figure 2 compares FOM values for coatings with as-sprayed surface roughness and with a polished 1 mm surface finish. These data show that the average FOM difference between the as-sprayed and polished coupons, calculated according to ΔFOM = [(FOMas – FOMpol)/FOMas]×100 was ~40% (as = as sprayed; pol = polished). Effect of heat treatment: Figure 3 shows the change in FOM for different coating compositions following heat treatment for six hours at 600°C in air. Heat treatment increased the FOM for all compositions. During heating, two aspects of the coating surface expected to change for all compositions are an increase in thickness of the surface oxide covering the coating and minimization of surface energy promotes diffusion, which reduces surface roughness. Spreading and solidification of liquid droplets should not generate a significant amount of high aspect ratio surface asperities that would significantly change shape with a postspraying heat treatment. Furthermore,

Properties of spinel coatings Spinels are oxide materials with the general formula AB2O4. A variety of stoichiometric spinel films, AB2O4 (A, B = Ni, Co, Fe, Cu), were formulated via dip and spin coating. Spinels were investigated as solar selective materials because of their inherent high temperature and oxidation stability [3-6]. They are also amenable to cation doping and substitution on both the A and B sites, which can affect their optical properties. Optical properties for Co3O4 are shown in Table 5. Increasing the film thickness (# Coatings, Table 5) leads to an increase in absorptance, but it also leads to an increase in emittance. Each Co3O4 film was aged at 500°C in air atmosphere for four days. Absorptance and emittance values were essentially the same before and after aging, suggesting that the films exhibit good thermal stability. The optical properties of 5 and 20wt% metal-doped Co3O4 are shown in Table 6. Neither doping concentration nor film thickness had an effect on coating absorptance.

35 25 15

Tungsten_HT

Tungsten

Ni-5Al

Ni-5Al_HT

5 Ni-25Graphite_HT

evidence presented above suggests decreasing the surface roughness produces a decrease in the FOM. Therefore, changes in the oxide layer on the coating surface are likely the dominant factor causing the FOM to change with heat treatment. The WC-Co coatings delaminated and fractured during heat treatment due to residual stress and/or coefficient of thermal expansion mismatch between the coating and substrate. The damage to the WC-Co coatings made it impossible to collect absorption and emittance data after heat treatment. These data are supported by published reports on the use of nickel-aluminium and tungsten carbide-cobalt alloys as solar selective coatings. Santala and Sabol produced nickel + 50wt% aluminium coatings by roll bonding that exhibited absorptance values >0.9 and emittance values