Open Access
Issue
Wuhan Univ. J. Nat. Sci.
Volume 27, Number 1, March 2022
Page(s) 77 - 84
DOI https://doi.org/10.1051/wujns/2022271077
Published online 16 March 2022

© Wuhan University 2022

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

0 Introduction

Reverse osmosis (RO) membrane technology, superior in high efficiency and good treatment effect, is a common advanced treatment technology for landfill leachate. However, membrane fouling will appear on the RO membrane with the operation of the treatment system, leading to membrane blockage, increased operation pressure, reduced water production rate, and other problems[1]. The adsorbed pollutants on the membrane surface and in the membrane pores are the main reason for the reduction of membrane flux and the deterioration of treatment effect. The pollutants of RO membrane caused by leachate are organic matter mainly composed of humic acid, fulvic acid and protein-like substances, as well as inorganic matter and colloids such as calcium and magnesium[2-4]. The properties of humic acid and fulvic acid in the leachate are similar, and their main bodies are aromatic compounds. The molecular weight of fulvic acid is relatively small, generally below 1 000, while that of humic acid is relatively large, generally between 1 000 and 10 000. Since pollutants are difficult to remove, membrane fouling is often irreversible. The system no longer works when its operating pressure reaches the established value. In order to keep it running, the RO membrane needs to be cleaned and replaced, thus increasing operating costs and limiting the promotion of RO membrane technology.

Incineration and landfill are the mainstream ways of treating municipal solid waste (MSW) in China[5]. The waste entering incineration plant has a short fermentation and oxidation time, and the leachate produced has a high moisture content and contains a large amount of biodegradable organic matter (such as volatile fatty acids). Therefore, the leachate of incineration plant has high chemical oxygen demand (COD) concentration and good biodegradability. Besides, it generally takes a long time to landfill and biodegrade wastes (generally more than 3 years). The leachate from landfill contains a large amount of humus high in humification degree and difficult to degrade (such as fulvic acid and humic acid), so it contains a low-concentration COD but demonstrates poor biodegradability. The pollution of RO membrane is related to the type, nature and concentration of influent pollutants[6]. The fouling process and mechanism of RO membrane may be different due to the different water quality of the two leachate.

At present, RO membrane technology is widely used in leachate treatment in incineration plant and landfill. In order to solve the problem arising from RO membrane fouling during leachate treatment, many researches have been conducted[7,8]. However, most of the current studies only focus on the cleaning of RO membrane[9-11], while there are relatively few studies on the composition analysis of membrane scale and its cumulative process. And there is no clear understanding of the types and quantities of pollutants, and there is no report on the comparison of RO membrane scaling mechanism between incinerator and landfill leachate treatment.

In this study, on the basis of three-dimensional excitation and emission matrix fluorescence (3DEEM) and nano-particle size comparative analysis of influent, effluent, concentrated liquid pollutants and membrane fouling from the disc tube reverse osmosis (DTRO) in typical incineration plants and landfill in Chongqing, a city in the west of China, the main components of membrane fouling were identified. Then, by discussing the change law of COD and main pollutants in the inner membrane pollution of DTRO system during a single operation cycle, the accumulation process of RO membrane pollutants was obtained, which provided a reference for understanding the composition, formation and control of membrane pollution.

1 Materials and Methods

1.1 Experimental Materials

The samples used in this experiment were taken from a typical incineration plant and a landfill in Chongqing, namely Fengsheng Incineration plant and Changshengqiao landfill leachate treatment system. The specific process flow and sampling point are shown in Fig. 1. The RO membrane component was cleaned once when the water inflow was 1 200 m3 and 1 000 m3, respectively.

thumbnail Fig. 1 Leachate treatment process and sampling points in incinerator (a) and landfill (b)

The influent, effluent and concentrated liquid of the DTRO system of the incineration plant were 100, 300, 500, 700, 900 and 1 100 m3. The influent, effluent and concentrated liquid produced by the first-stage RO system in the landfill were 100, 300, 500, 700 and 900 m3.

Dow RO membranes obtained in the treatment process after one cycle of DTRO system operation and before reverse cleaning were used for incineration plant and landfill.

The instruments are listed in Table 1.

Table1

Information and use of laboratory equipment

1.2 Determination of RO Membrane Fouling Composition

3DEEM characterizes the fluorescence peaks of organic matter in leachate with different emission and excitation wavelengths, and provides complete spectral information, thus revealing the source and component structure of different organic pollutants[12-14]. However, as the fluorescence peaks are usually signals generated by multiple superimposed fluorescence groups, some of them may be unrecognizable or inaccurate. Fluorescence regional integration (FRI), another fluorescence spectrum analysis method, can identify and characterize the overlapping objects of fluorescence spectra in multi- component systems, and overcome the above deficiencies to a certain extent[15]. Some studies[16-19] divide 3DEEM spectrum into five regions, as shown in Table 2. The fluorescence signal of region Ⅲ is mainly generated by substances like fulvic acid in ultraviolet region, while region Ⅴ is related to organics with higher molecular weight and aromatization degree such as fulvic acid, humic acid and polycyclic aromatic hydrocarbon in visible region.

On the basis of unit fluorescence spectrum, each region divided by 3DEEM was calculated quantitatively[20] by FRI. The discrete integral formula is expressed as:(1)where is integral value, excitation wavelength interval (take 5 nm), emission wavelength interval (take 5 nm), and fluorescence intensity (FI) at each point.

Table 2

Zoning of 3DEEM

1.3 Determination of RO Membrane Fouling Formation Process

The indicators in line with the Pollution Control Standard of Domestic Waste Landfill (GBL6889-2008) were used for monitoring in this experiment. The high content of COD in landfill leachate represents a large amount of organic matter and other pollutants in the leachate, and is a potential contaminant of RO membrane, which is clearly required by the discharge standard. Therefore, COD concentration is one of the important detection indexes for RO membrane pollution process analysis.

1.3.1 Conservation of materials

According to the conservation of material, the content of membrane fouling can be calculated by that of inlet water, outlet water and concentrated liquid.(2)(3)(4)(5)where C, V, n, d, and c represent the concentration, volume, amount of substance, number of membranes and number of membrane columns, respectively. “TI” and “IN” represent the total influent and influent, “TE” and “EF” represent the total effluent and effluent, “TC” and “CL” represent the total concentrated liquid and concentrated liquid, “TD” and “DI” represent the total dirt and dirt. The coefficient λ represents the ratio of material output to input of the DTRO system:(6)

It is assumed that the water quality of 200 m3 leachate near each water sample remained unchanged, and the water production rate η of DTRO system in incineration plant and landfill was 75%. At the end of DTRO operation, three RO films with fouling were taken, and the fouling was dissolved in 100 mL ultra-pure water, respectively. COD of the fouling solution was then measured, and the average value was taken. The coefficient λ corresponding to COD was obtained from formulas (2) to (6) to calculate the conservation of materials.

1.3.2 Linear relationship between FI and fulvic acid

According to Shi’s study[21], the concentration of organic compounds can be calculated by the FI of each component of leachate. In addition, Liu and Wu[22] indicated that the concentration of fulvic acid in water samples can be accurately reflected by FI analysis of 3DEEM spectrum. In order to quantitatively analyze the total amount of fulvic acid in leachate, this study set a voltage to be 700 V and used FI under =320 nm/400 nm as 3DEEM line of fulvic acid[23]. The content of fulvic acid in water sample was calculated by measuring peak height. According to the standard curve of fulvic acid and the conservation of materials, the content of fulvic acid in membrane scale was calculated, and the amount of scale formation and the total amount of accumulated scale formation of leachate influent per 200 m3 in the DTRO system of incineration plant and landfill was further obtained.

2 Results and Discussion

2.1 Component Analysis of Membrane Fouling

As can be seen from Fig. 2, seven fluorescence characteristic peaks appeared in the water sample during the experiment, which were visible fulvic acid fluoresc- ence peak (peak A), UV fulvic acid fluorescence peak (peak B), aromatic protein (peak C), highly excited tryptophan protein (peak D), highly excited tyrosine protein (peak E), low-excitation tyrosine protein (peak F), and low-excitation tryptophan protein (peak G). The 3DEEM isolines of RO fouling solution in the incineration plant and landfill were sparsely distributed, indicating the relatively small content of organic matter in fouling. Combined with Fig. 3, it can be seen that no matter in incineration plant or landfill, the proportion of organic matter in zones Ⅲ and Ⅴ was the largest, which was 51.52% and 33.25% for incineration plant, and 34.13% and 55.26% for landfill, respectively. The main pollutants in zones Ⅲ and Ⅴ included UV fulvic acid and visible fulvic acid respectively. Therefore, the main pollutants in RO membrane fouling could be UV fulvic acid and visible fulvic acid, which is consistent with the research results obtained by Han[24]. UV fulvic acid of incineration plant and UV fulvic acid of landfill accounted for more than half of their respective organic pollutants.

thumbnail Fig. 2 3DEEM of RO membrane fouling

thumbnail Fig. 3 Proportion of RO membrane fouling in organic matter integral value of each region on 3DEEM

According to Table 3, the peak value of peak B (75.9) was significantly higher than that of peak A (51.8), indicating that the UV fulvic acid with simple structure and small molecular weight possibly blocked the pores on the membrane surface and thus these pores gradually accumulated and deposited on the membrane surface, resulting in poor membrane permeability and membrane pollution. This is consistent with Meng’s research[25] statement. Peaks D, F and G were prominent, and the fluorescence peaks were higher than peaks A and B, indicating that the membrane fouling respectively contained more protein substances and more high-excitation tryptophan with low-excitation tyrosine and low- excitation tryptophan. The fluorescence peak of peak E was as high as 213.2, which is clearly seen on the spectrum, indicating that the highly excited tyrosine protein may also be one of the main contaminants of the membrane. According to the study by Li et al[26], protein-like substances have an important effect on membrane fouling. The higher the concentration of protein-like substances, the greater the possibility of membrane fouling. It contains more protein and humic acid substances. Membrane fouling is more likely to occur, which is consistent with the above research conclusions.

As can be concluded from the landfill data, both peak A and peak B had a strong response. The peak A (651.5) was significantly higher than the peak B, possibly due to the fact that the accumulated UV fulvic acid resulted in film pollution, and that the molecular weight was relatively large. Visible fulvic acid was retained on the RO membrane. Peaks D, E and F were significantly lower than peaks A and B, indicating that the lower content of high-excitation tryptophan, high-excitation tyrosine and low-excitation tyrosine in the dirt. The peak value of peak G was as high as 205.5, and there was a strong response, indicating that low-excitation tryptophan-like proteins may be one of the main pollutants of RO membranes.

According to Zhang et al[27], tyrosine-based proteins and tryptophan-based proteins are membrane contaminants that are difficult to remove. Therefore, combined with the above analysis, the organic matter difficult to remove in the RO membrane treatment system of incineration plants and landfills may be highly excited tyrosine protein and low-excitation tryptophan protein, respectively.

According to the study by Zhang et al[27], FRI can reflect the relative content of organic matter represented by each region. Cui et al[28] showed that r(B/A),the ratio of fluorescence intensity between peak B and peak A, is an index related to the maturity and structure of organic matter, and can be used to represent the degree of humification of dissolved organic matter (DOM) in leachate. The higher the r(B/A) value, the lower the degree of humification of leachate. The r(G/E) value indicates the composition and degradation of proteins in leachate. The higher the r(G/E) value, the more complex the proteins in leachate, the worse the biodegradability. The r(G/B) value represents the proportion of protein substances in leachate. The higher the r(G/B) value, the larger the proportion of protein in leachate.

As can be seen from Table 4, DOM humification in RO membrane fouling in landfill was greater than that in incineration plant, and protein composition was more complex than that in incineration plant. However, the proportion of DOM protein in RO membrane fouling in landfill was not as high as that in incinerator, which indicated that in the process of RO, proteins in the membrane fouling of the incineration plant continued to accumulate, becoming the main pollutant of membrane fouling, which is consistent with the analysis result of fluorescence integral.

According to Baun et al[29] and Lou et al[30], the pollutants in leachate can be divided into three parts according to particle size by membrane technology, namely, soluble substances and fine colloid (<450 nm), coarse colloid (450-1 200 nm), and suspended matter (>1 200 nm). The organic matter in leachate can be divided into alcohols, organic acids and fulvic acid of small molecular weight, as well as fulvic acid and protein high molecular weight[31-33].

As can be seen from Table 5, the particle sizes of pollutants in RO membrane fouling were mostly in the range of 450-1 200 nm, and the proportions of the incineration plant and landfill were 71.2% and 89.5%, respectively. This area was dominated by fulvic acid with small molecular weight. The above analysis results show that fulvic acid is the main pollutant in membrane fouling. Therefore, the main pollutant of membrane fouling is small-molecule fulvic acid with particle size ranging from 450 nm to 1 200 nm. According to Zhang’s research[34], the fulvic acid with small molecular weight is mainly distributed in the gel body, making it difficult to penetrate the membrane, thus resulting in the continuous concentration of concentrated liquid and the increase of particle size. In addition, according to Wang et al[35], fulvic acid with small molecular weight is the main cause of RO membrane pollution.

Table 3

Peak position and intensity of fluorescent substances in RO membrane fouling

Table 4

FI ratio of DOM in RO membrane fouling

Table 5

Distribution of particle size of membrane fouling

2.2 Accumulation Rule of Membrane Fouling and Fulvic Acid

The COD coefficient λ of the incineration plant and the landfill DTRO system was calculated to be close to 100%, so the COD in the membrane fouling can be calculated by material conservation. Combined with Fig. 4 and Fig. 5, it can be seen that the scaling law of RO membrane in incineration plant and landfill was similar. With the operation of DTRO system, the amount of COD and fulvic acid in membrane fouling per unit of water inflow exhibited a downward trend, indicating the large amount of scale per unit in the first half of the operation cycle (the first half of the water inflow stage) possibly because of the smooth membrane surface in the first half of operation and the high likelihood of pollutants to adsorb and deposit on the membrane surface[36]. In the later stage of operation, the membrane surface was rough, and the amount of pollutants and pressure increases, which made it quite difficult for these pollutants to adhere to the membrane surface. In the same operation cycle, COD and fulvic acid per unit of influent water in the fouling of RO membrane in landfill were significantly higher than those in incinerator, indicating that the RO membrane in landfill was easier to scale than that in incinerator.

thumbnail Fig. 4 COD of membrane fouling in one operation cycle of DTRO system

thumbnail Fig. 5 Change of fulvic acid in membrane fouling during one operation cycle of DTRO system

3 Conclusion

The main pollutants in RO membrane fouling of the incineration plant and landfill are UV fulvic acid and visible fulvic acid with particle size ranging from 450 nm to 1 200 nm, and the pollutants that are difficult to remove include highly excited tyrosine protein and low-excitation tryptophan protein, respectively.

The membrane pollution of RO membrane in incineration plant and landfill mainly occurs in the first half of system operation. With the operation of DTRO system, the amount of COD and fulvic acid in membrane fouling per unit influent shows a downward trend. In the same operation period, the RO membrane in landfill is more likely to scale than that in incineration plant.

References

  1. Yang Z G. Process and Principal of Membrane Science and Technology [M]. Shanghai: East China University of Science and Technology Press, 2009(Ch). [Google Scholar]
  2. Zhao S F, Liao Z P, Fane A, et al. Engineering antifouling reverse osmosis membranes: A review [J]. Desalination, 2021, 499: 114857. [Google Scholar]
  3. Liu Y P, Li X J. Study on RO membrane fouling in landfill leachate treatment [J]. Journal of Environmental Engineering, 2007, 1(7): 101-105(Ch). [Google Scholar]
  4. Lou M L. Study on Membrane Fouling by Typical Pollutants and Membrane Cleaning [D]. Tianjin: Tianjin Institute of Urban Construction, 2010(Ch). [Google Scholar]
  5. Cheng W, Quan X J, Huang X X, et al. Enhancement of micro-filtration performance for biologically-treated leachate from municipal solid waste by ozonation in a micro bubble reactor [J]. Separation and Purification Technology, 2018, 207: 535-542. [CrossRef] [Google Scholar]
  6. Xiao S Q. Reverse osmosis membrane fouling mechanism and prevention and control measures [J]. Guangdong Chemical Industry, 2016, 43(21): 119-120+123(Ch). [Google Scholar]
  7. Luo M L, Li L, Jiang R. Research on the pollution of RO membrane by BSA [J]. Industrial Water Treatment, 2018, 38(4): 53-56(Ch). [Google Scholar]
  8. Ren X, Song K, Xiao Y, et al. Effective treatment of spacer tube reverse osmosis membrane concentrated leachate from an incineration power plant using coagulation coupled with electrochemical treatment processes [J]. Chemosphere, 2020, 244: 125479. [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  9. Zhou D Y, Wang W, Qi Q. Analysis on reverse osmosis membrane fouling and its on-line cleaning [J]. Cleaning World, 2016, 32(7): 22-26(Ch). [Google Scholar]
  10. Zhao B. Analysis of reverse osmosis membrane chemical cleaning [J]. Chemical Engineering Design Communications, 2020, 46(10): 2(Ch). [Google Scholar]
  11. Baatiyyah H. The Impact of Chemical Cleaning on Separation Efficiency and Properties of Reverse Osmosis Membrane [D]. Jeddah: King Abdullah University of Science and Technology, 2018. [Google Scholar]
  12. Peng N, Wang K F, Liu G G, et al. Quantifying interactions between propranolol and dissolved organic matter (DOM) from different sources using fluorescence spectroscopy [J]. Environmental Science and Pollution Research, 2014, 21(7), 5217-5226. [CrossRef] [PubMed] [Google Scholar]
  13. Ouyang E M, Zhang X H, Wang W. Study on the effects of organic removal by traditional purification process with three-dimensional excitation emission matrix fluorescence spectroscopy [J]. Spectroscopy and Spectral Analysis, 2007, 27(7): 1373-1376(Ch). [Google Scholar]
  14. Lv H G, Ouyang E M, Zheng Z H, et al. Three dimensional fluorescence spectroscopy for water quality determination in water treatment plant [J]. China Water & Wastewater, 2005, 21(3): 91-93(Ch). [Google Scholar]
  15. Guo W D, Huang J P, Hong H S, et al. Resolving excitation emission matrix spectroscopy of estuarine CDOM with parallel factor analysis and its application in organic pollution monitoring [J]. Environmental Science, 2010, 31(6): 1419-1427(Ch). [Google Scholar]
  16. Wu J, Cui S, Xie C B, et al. Fluorescence fingerprint transformation of municipal wastewater caused by aerobic treatment [J]. Spectroscopy and Spectral Analysis, 2011, 31(12): 3302-3306(Ch). [Google Scholar]
  17. Peiris R H, Budman H, Moresoli C, et al. Understanding fouling behaviour of ultrafiltration membrane processes and natural water using principal component analysis of fluorescence excitation-emission matrices [J]. Journal of Membrane Science, 2010, 357(1-2): 62-72. [CrossRef] [Google Scholar]
  18. Zhu H T, Wen X H, Huang X. Membrane organic fouling and the effect of pre-ozonation in microfiltration of secondary effluent organic matter [J]. Journal of Membrane Science, 2010, 352(1-2): 213-221. [NASA ADS] [CrossRef] [Google Scholar]
  19. Fan L, Nguyen T, Roddick F A, et al. Low-pressure membrane filtration of secondary effluent in water reuse: Pre- treatment for fouling reduction [J]. Journal of Membrane Science, 2008, 320(1-2): 135-142. [CrossRef] [Google Scholar]
  20. Han R B. Removal of Micro-polluted Organic Matter from Source Water by Membrane Bioreactor Combined with Powdered Activated Carbon [D]. Harbin: Harbin Institute of Technology, 2013(Ch). [Google Scholar]
  21. Shi J. Based on the 3D Fluorescence Spectra of Sewage Treatment Plant Water Quality Quick Analysis [D]. Yangzhou: Yangzhou University, 2011(Ch). [Google Scholar]
  22. Liu W, Wu Y F. Fluorescence spectrum analysis of photocatalytic degradation of fulvic acid [C]// The 2010 Annual Meeting of the National Water Supply Advanced Treatment Research Association of the Water Industry Branch of the Chinese Civil Engineering Society. Shanghai: The Water Industry Branch of the Chinese Civil Engineering Society, 2010: 494- 501(Ch). [Google Scholar]
  23. Zhao Q L, Jia T, Wei L L, et al. Fluorescence characterization of the reactivity of THMs precursors in secondary effluent [J]. China Environmental Science, 2009, 29(11): 1164-1170(Ch). [Google Scholar]
  24. Han L. Study on Fouling Mechanism and Control Technology of Leachate Treated by Reverse Osmosis Membrane [D]. Chongqing: Chongqing University, 2012(Ch). [Google Scholar]
  25. Meng F, Chae S R, Drews A, et al. Recent advances in membrane bioreactors (MBRs): Membrane fouling and membrane material [J]. Water Research, 2009, 43(6): 1489-1512. [CrossRef] [PubMed] [Google Scholar]
  26. Li S, Duan L, Zhou B H, et al. Effect of soluble microbial products on membrane fouling in MBR at various solids retention times [J]. Chinese Journal of Environmental Engineering, 2015, 9(6): 2731-2738(Ch). [Google Scholar]
  27. Zhang Y J, Sun L H, Chen X R, et al. Adsorption of powdered activated carbon and activated coke towards organics removal by ultra-filtration in reclaimed wastewater treatment [J]. Chinese Journal of Environmental Engineering, 2015, 9(7): 3099-3105(Ch). [Google Scholar]
  28. Cui H R, Cai X M, Qin X, et al. Hydrogen peroxide catalytic wet oxidation of leachate and the spectroscopic analysis of dissolved organic matters [J]. Acta Scientiae Circumstantiae, 2015, 35(9): 2930-2935. [Google Scholar]
  29. Baun D L, Christensen T H. Speciation of heavy metals in landfill leachate: A review [J]. Waste Management & Research, 2004, 22(1): 3-23. [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  30. Lou Z Y, Zhao Y C. Gradient size-fractionation and characterization of leachate by micro-membrane filter [J]. Nonferrous Metals Engineering & Research, 2009, 30(6): 69-71. [Google Scholar]
  31. Liu X G. Study on New Type of Polyacrylamide Modified Bentonite Impermeable Materials [D]. Shenyang: Northeastern University, 2010(Ch). [Google Scholar]
  32. Ding Z. Study on Component Characteristics of Landfill Leachate and Microwave Advanced Oxidation Treatment [D]. Xi’an: Chang’an University, 2009(Ch). [Google Scholar]
  33. Xu Y D, Yue D B, Nie Y F. The size distribution for pollutants in landfill leachate and its implications [J]. Techniques and Equipment for Environmental Pollution Control, 2006, 7(5): 16-21(Ch). [Google Scholar]
  34. Zhang Y F. Study on Water Quality of Leachate from Garbage Landfill Site and Garbage Incineration Plant in Chongqing’s Main City Zone [D]. Chongqing: Southwest University, 2008(Ch). [Google Scholar]
  35. Wang Y, Li B, Dong Z Y, et al. Gradient separation for pollutants in leachate of a sanitary landfill and its characterization [J]. Chinese Journal of Environmental Engineering, 2012, 6(4): 1281-1287(Ch). [Google Scholar]
  36. Chu Y J, Yu H Q, Cui C. Study on contaminants and their distribution on the surface of RO membrane [J]. Technology of Water Treatment, 2012, 38(1): 72-74(Ch). [Google Scholar]

All Tables

Table1

Information and use of laboratory equipment

Table 2

Zoning of 3DEEM

Table 3

Peak position and intensity of fluorescent substances in RO membrane fouling

Table 4

FI ratio of DOM in RO membrane fouling

Table 5

Distribution of particle size of membrane fouling

All Figures

thumbnail Fig. 1 Leachate treatment process and sampling points in incinerator (a) and landfill (b)
In the text
thumbnail Fig. 2 3DEEM of RO membrane fouling
In the text
thumbnail Fig. 3 Proportion of RO membrane fouling in organic matter integral value of each region on 3DEEM
In the text
thumbnail Fig. 4 COD of membrane fouling in one operation cycle of DTRO system
In the text
thumbnail Fig. 5 Change of fulvic acid in membrane fouling during one operation cycle of DTRO system
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.