8th Annual California Unified Program Conference

1
8th Annual California Unified Program Conference
Advanced Hazardous Waste Inspector Training
2
What is a valid Waste Determination?

• Part II. Analysis or
• Knowledge of Process?

3
Most of the Time

• Its simple

4
But when it isn't simple, who makes the waste
determination?

• The Generator
• The person whose act or process produces
  hazardous waste or whose act first causes a
  hazardous waste to become subject to regulation.
• A hazardous waste Generator must comply with the
  requirements of Title 22 CCR, Division 4.5,
  Chapter 12.

5
66262.11 Hazardous Waste Determination

• First, the generator must determine if it is a
  waste.
• 66261.2 Definition of a waste
• 66261.3 Definition of a hazardous waste
• 66261.4 Materials which are not waste
• 25143.2 Excluded recyclable materials
• Next, the generator must determine if it is a
  hazardous waste.
• Is it listed in article 4 or in Appendix X of
  Chapter 11?
• Or does it exhibit any of the characteristics set
  forth in article 3 of Chapter 11?

6
66262.11 Hazardous Waste Determination (cont).

• The Generator can make a hazardous waste
  determination by
• (1) Testing or
• (2) Applying knowledge of the hazard
  characteristic of the waste in light of the
  materials or the processes used.
• This is also called waste analysis.

7
Waste Analysis

• The cornerstone of a hazardous waste program is
  the ability of facility personnel to identify
  properly, through waste analysis, all the wastes
  they generate, treat, store, or dispose of.
  Waste analysis involves identifying or verifying
  the chemical and physical characteristics of a
  waste by performing a detailed chemical and
  physical analysis of a representative sample of
  the waste, or in certain cases, by applying
  acceptable knowledge of the waste.

8
Testing

• Accurate analytical data is required to comply
  with Chapter 18, LDR requirements.
• A written Waste Analysis Plan (WAP) is required
  for
• TSDFs,
• PBR Treatment and,
• Generators treating hazardous to meet LDR
  standards.

9
A Waste Analysis Plan

• Establishes consistent internal management
  mechanism(s) for properly identifying wastes on
  site.
• Ensures that waste analysis participants have
  identical information (e.g., a hands-on operating
  manual), promoting consistency and decreasing
  errors.
• Ensures that facility personnel changes or
  absences do not lead to lost information.
• Reduces your liabilities by decreasing the
  instances of improper handling or management of
  wastes.

10
Waste Analysis Plan?

• http//www.epa.gov/epaoswer/hazwaste/ldr/wap330.pd
  f

11
Article 3. Characteristics of Hazardous
Waste66261.20.General

• (a) A waste, as defined in section 66261.2, which
  is not excluded from regulation as a hazardous
  waste pursuant to section 66261.4(b), is a
  hazardous waste if it exhibits any of the
  characteristics identified in this article
• (c) Sampling and testing pursuant to this article
  shall be in accord with the chapter nine of
  SW-846, the Department will consider samples
  obtained using any of the other applicable
  sampling methods specified in Appendix I of this
  chapter to be representative samples.

12
Characteristic Wastes

• 66261.21 (a) A waste exhibits the characteristic
  of ignitability if representative samples of the
  waste have any of the following properties
• 66261.22(a) A waste exhibits the characteristic
  of corrosivity if representative samples of the
  waste have any of the following properties
• 66261.23(a) A waste exhibits the characteristic
  of reactivity if representative samples of the
  waste have any of the following properties
• 66261.24 (a) A waste exhibits the characteristic
  of toxicity if representative samples of the
  waste have any of the following properties

13
Representative Sample

• 66260.10. Definitions. "Representative sample"
  means a sample of a universe or whole (e.g.,
  waste pile, lagoon, ground water) which can be
  expected to exhibit the average properties of the
  universe or whole.

14
NoteEnforcement Sample

• A regulator does not necessarily need a
  representative sample to support an enforcement
  action.
• The primary reason is that the data quality
  objectives (DQOs) of the enforcement agency often
  may be legitimately different from those of a
  waste handler.
• A sample taken for enforcement is used to
  demonstrate that the waste exceeds a standard
  (e.g. STLC).

15
EPA publication SW-846 Test Methods for
Evaluating Solid Waste, Physical/Chemical Methods

• OSW's official compendium of approved analytical
  and sampling methods for use in complying with
  RCRA regulations.
• SW-846 primarily is a guidance document that sets
  forth acceptable, although not required, methods
  for the regulated and regulatory communities to
  use for RCRA-related sampling and analysis
  requirements.

16
SW 846

• http//www.epa.gov/sw-846/sw846.htm

17
SW 846, Chapter 9,Sampling Plan

• SW 846 assumes that
• The concentration of a contaminant in individual
  samples will exhibit a normal distribution.
• Simple random sampling is the most appropriate
  sampling strategy.
• As more information is accumulated, greater
  consideration can be given to different sampling
  strategies.
• Start with simple random sampling and assume a
  normal distribution.

18
Population

• 90 bags of candy
• 10 bags contain 0 pieces of (0 pieces)
• 20 bags contain 1 piece of (20 pieces)
• 30 bags contain 2 pieces of (60 pieces)
• 20 bags contain 3 pieces of (60 pieces)
• 10 bags contain 4 pieces of (40 pieces)
• Population mean is 180/90 2

19
Histograph of candy population (normal
distribution)
Population mean 2
30 20 10
0 1 2 3 4
20
Random sample

• Four samples from web 14, 37, 40, 81 (90 bags)
• Four samples from web 7, 19, 35, 41 (50 bags)
• Six samples from web 3, 24, 64, 71, 76 , 90

21
Histograph of Samples
3 2 1
0 1 2 3 4
22
Normal Distribution
In a normal distribution a bell shaped curve is
used to represent the boundaries of the
population. The true population (under the blue
curve) is never known, but precise and unbiased
samples will provide an accurate estimate of the
true population.
Samples
The sample population under the magenta curve is
an estimate of the true population.
23
A Bell Curve has Tails!
I left the tails off most of the diagrams because
I couldnt figure out how to draw them!

• The X axis is the concentration.
• The Y axis is the number of samples.
• The tails are where the people who
• got 100 or 0 on an exam are found.

24
Reliable Waste Analysis

• Reliable information concerning the chemical
  properties of a solid waste is needed for
  comparison with applicable regulatory thresholds.
  
• If chemical information is to be considered
  reliable, it must be accurate and sufficiently
  precise.
• Accuracy (no bias) is usually achieved by
  incorporating randomness into the sample
  selection process.
• Sufficient precision is most often obtained by
  selecting an appropriate number of samples.

25
Sample size

• Small samples (A) cause the constituent of
  interest to be under-represented in most samples
  and over-represented in a small proportion of
  samples. Larger samples (B) more closely reflect
  the parent population.
• Sometimes you sample a large portion or even the
  entire population, so you dont need statistics
  to determine a confidence interval.

26
TerminologyPrecise, Accurate Biased

• Precise means all of the samples are similar
  they form a tight group on the graph. Taking
  more samples or taking larger samples will
  increase the sample precision.
• Accurate or unbiased means that youre taking
  truly random samples. Properly planned random
  samples are accurate and unbiased samples.
• Inaccurate samples are synonymous with biased
  samples. They are not representative samples.
  Poor tool selection or calibration can cause
  sample bias.

27
Biased Imprecise Samples
Biased samples do not represent the true
population. The biases could result from poor
tool selection or contamination.
Imprecise samples have a lot of variation. More
samples should decrease variation.
Mean 1012.5
0
2000
28
Biased Precise Samples
A poor sampling plan could lead to biased or
inaccurate samples. Poor tool selection, poor
sampling design or contamination are some causes.
Biased sampling shifts the population curve.
Sample Mean
True Mean
0
2000
Who can think of another cause for biased samples?
29
Unbiased Imprecise Samples
Unbiased samples are Random samples. Random
samples fall inside the bell curve that
represents the true population.
Take more samples to increase the precision.
Mean 1012.5
0
2000
30
Unbiased Precise Samples
Unbiased samples are a function of randomness.
Random sampling requires proper plan design and
tool selection.
Precise samples are a function of the number of
samples.
Mean 1012.5
0
2000
31
Waste Analysis (Testing) To evaluate the
physical and chemical properties of a solid waste

• The initial -- and perhaps most critical --
  element is the sampling plan.
• Analytical studies, with their sophisticated
  instrumentation and high cost, are often
  perceived as the dominant element.
• But analytical data generated by a scientifically
  defective sampling plan have limited utility.

32
SW 846

• Waste characterization requires a representative
  sample.
• At least two samples of a material are required
  for any estimate of precision.
• SW 846 uses an 80 confidence interval as an
  acceptable degree of sampling accuracy and
  precision.
• Normally data from four representative samples is
  the minimum required to achieve an 80 confidence
  interval.

33
How many samples are enough? An example

• A business wants to dispose of a pile of used
  blast medium. It has been reused and it is well
  mixed. It might have been used to remove paint
  with lead pigment.
• Is it hazardous?
• Testing or knowledge of process?
• It might have lead? Knowledge??
• How many samples do need for testing?
• Four?

34
Sampling Plan

• Make a 3-D grid of the pile. Number each area of
  the grid.
• Select four numbers randomly. Random number
  generators are on the web, tables or in
  textbooks.
• Sample from the four areas represented by the
  number.
• Analyze the samples using TTLC.

35
Sample Results

• The TTLC for lead is 1000 mg/kg.
• Sample A contains 1000 mg/kg. Is sample A
  hazardous waste?
• Is the waste pile hazardous?
• Sample B contains 1050 mg/kg, sample C contains
  980 mg/kg and sample D contains 1020 mg/kg.
• Is the waste pile hazardous?

36
Is it hazardous?

• Yes, 3 of 4 is good enough.
• No, its 100 or nothing.
• More analysis and maybe more samples are required.

The answer is C!
37
More Analysis?

• Yes, more analysis.
• The samples were pretty close, A contains 1000
  mg/kg, B contains 1050, C 980 D 1020.
• A range of only 70 mg/kg.
• Do we need more samples?
• Yes, well

38
Guess how many samples

• 4

5
15
20?
The answer is 15.31
Where did that number come from?
39
A Seven step Statistical Process is used to
determine number of samples (SW 846 Table 9-1)

• Determine the mean
• Determine the variance
• Determine the standard deviation
• Determine the standard error
• Determine the confidence interval
• Determine if the variance is the mean
• Determine the appropriate number of samples.

40
Statistics, the last time

• I would have gotten a PHD if I liked math.
• Give it a chance!
• Its just addition, multiplication and division.
• Oh, and square roots, but you can use a
  calculator.

41
If you really hate Numbers

• Pretend to listen, its the
• polite thing to do, and
• remember
• You need at least four (4) samples.
• More samples may be required if the waste is
• Heterogeneous, or
• Close to the regulatory threshold

42
Step 1 The Mean
Samples A 1000ppm B 1050ppm C 980ppm D
1020ppm
The sample mean is the average value of the
samples. Its an estimate. The true mean is never
known.
Sample Mean 1012.5
0
2000
43
Normal DistributionVariance
The variance is the sum of the differences
between the sample values and the mean, squared.
?variance?
Mean
0
2000
The variance sets the boundaries of the
distribution.
44
Standard Deviation
Standard Deviation
The standard deviation is the square root of
the variance.
?variance?
Mean 1012.5
0
2000
45
Normal Distribution CI
80 Confidence Interval (CI)
If you take 100 samples, 80 should fall inside
the boundaries of the 80 CI.
?variance?
Mean 1012.5
0
2000
46
Normally you would evaluate all four samples

• All four randomly selected samples must be
  considered in a valid statistical analysis.
• In the following example, four sets of two will
  also be analyzed to illustrate the effects of
• Decreasing the variance in concentration in the
  samples.
• Increasing number of samples.
• The relationship of the mean to the Regulatory
  Threshold (RT).

47
Step 1. The Mean

• Add the results of all samples and divide by the
  number of samples
• Sample A1000ppm Sample B1050ppm
• Sample C980ppm Sample D1020ppm
• MEAN
• ABCD (100010509801020)/4 4050/4 1012.5
  ppm
• A B 2050/2 1025 ppm
• C D 2000/2 1000 ppm
• B D 2070/2 1035 ppm
• A C 1980/2 990 ppm

48
Step 2. The Variance

• Variance (sample A - mean)2 (sample B -
  mean)2 (..) Number of samples - 1
• (1000-1012.5)2(1050-1012.5) 2 (980-1012.5) 2
  (1020-1012.5)2
• 3
• (12.5)2 (37.5)2 (32.5)2 (7.5)2 2675/3
  891.67
• 3

AB (1000-1025)2 (1050-1025) 2
1250 1 CD ( 980 - 1000) 2 (1020
- 1000) 2 800 1 BD (1050 - 1035) 2
(1020 - 1035) 2 450 1 AC (1000 -
990) 2 ( 980 - 990) 2 200 1
49
Step 3 Standard Deviation

• A1000 ppm, B1050 ppm, C 980 ppm, D1020 ppm
• Standard Deviation Variance 1/2
• The variance of ABCD is 891.67 the square
  root of 891.67 (standard deviation) 29.86
• AB Variance 1250 standard deviation 35.35
• CD Variance 800 standard deviation 28.28
• BD Variance 450 standard deviation 21.21
• AC Variance 200 standard deviation 14.14

50
Step 4 Standard Error

• A1000ppm, B1050ppm, C 980ppm, D1020 ppm
• Standard Error Standard Deviation
• (Number of samples) ½
• Standard error ABCD 29.86/(4)1/2 14.93
• Standard error A B 35.35/1.41 25.07
• Standard error C D 28.28/1.41 20.06
• Standard error B D 21.21/1.41 15.04
• Standard error A C 14.14/1.41 10.03

51
Step 5Confidence Interval

• A1000ppm, B1050ppm, C 980ppm, D1020 ppm
• Confidence Interval Mean (student
  t)(standard error)
• ABCD 1012.5 (1.638)(14.93) 1012.5 25.46
  (988 to 1038). 80 of 100 samples should have
  concentrations between 988 and 1038 ppm.
• AB 1025 (3.078)(25.07) 1025 77 (948 to
  1102)
• CD 1000 (3.078)(20.06) 1000 62 (938 to
  1062)
• BD 1035 (3.078)(15.04) 1035 46 (989 to
  1081)
• AC 1010 (3.078)(10.03) 1010 31 (979 to
  1041)

52
Step 6. Is the Variancethe Mean?

• If the variance is not greater than the mean, go
  to step 7.
• A B C D 891.67 is not 1012.5
• If the variance is greater than the mean , you
  have to transform the data. An example follows
  for samples A B.
• AB 1250 is 1025
• CD 800 is not 1000
• BD 450 is not 1035
• AC 200 is not 990

53
Is the Variance the Mean?
Mean
0
0
Variance
If variance is mean then part of the population
is less than zero, i.e. with samples A B the
population is between -225 and 2275. You cant
have a concentration of less than zero so you
have to transform the data.
54
Not more math!

• OK, we wont transform the data, here
• But in your handout the next four slides take the
  square root and go through the steps 1 to 6 and
  square the data to return to real numbers.
• Go to step 7.

55
Transform the data if the variance is the mean

• Usually data is transformed into a smaller number
  by taking either the log or the square root of
  the value.
• Step 1a. Transforming the mean
• 10001/2 31.62
• 10501/2 32.40
• Total 64.02/2 32.01

56
Transforming the Variance Standard Deviation

• Step 2a. Transforming the Variance
• Variance (sample A - mean)2 (sample B -
  mean)2
• Number of samples - 1
• AB (31.62 32.01)2 (32.4 - 32.01) 2 0.304
• 1
• Step 3a. Transforming the Standard Deviation
• Standard Deviation Variance 1/2
• AB (0.304)1/2 0.5515

57
Transforming the Standard Error and Confidence
Interval

• Step 4a. Transforming the Standard Error
• Standard Error Standard Deviation
• (Number of samples) 1/2
• AB 0.5515/1.41 0.39
• Step 5a. Transforming the Confidence Interval
• Mean (student t)(standard error)
• AB 32.01 (3.078)(.39) 32.01 1.20

58
Step 6aVariance Mean

• AB The transformed variance (0.304) is not
  greater than the transformed mean (32.01).
• Now go to the last step 7.

59
Step 7. Determine the number of samples (n)

• The Regulatory Threshold (RT) using TTLC for lead
  is 1000 ppm.
• n (student t)2(variance)
• (RT mean)2d
• Use the square root of the RT for lead (36.62)
• for transformed data.

60
n (student t)2(variance)(RT mean)2

• A B C D (1.638)2 (892) 15.31
• (1000 1012.5)2
• AB (3.078)2 (0.304) 7.73 samples
• (32.62 32.01)2
• CD (3.078)2 (800) ?
• (1000 1000)2
• BD (3.078)2 (450) 4.27
• (1000 1035)2
• AC (3.078)2 (200) 18.94
• (1000 990)2

61
So, fewer samples are required if,

• The waste is essentially homogenous
• or
• Well above or below the threshold

62
Other Types of Sampling

• Stratified random sampling
• Systematic random sampling
• Authoritative sampling

63
Stratified random sampling

• Stratified random sampling is appropriate if a
  batch of waste is known to be non-randomly
  heterogeneous.
• An example is a pile of blast media. One layer is
  from blasting lead paint, the next layer is from
  blasting new aluminum parts prior to painting.
  Another example is a stripping tank that is used
  to clean different parts and is periodically
  changed. The waste could vary from batch to
  batch.
• Stratification may occur over space (locations or
  points in a batch of waste) and/or time (each
  batch of waste).
• The units in each stratum are numerically
  identified, and a simple random sample is taken
  from each stratum.

64
Systematic random sampling

• Systematic random sampling, in which the first
  unit to be collected from a population is
  randomly selected but all subsequent units are
  taken at fixed space or time intervals.
• An example of systematic random sampling is the
  sampling along a pipeline at 20 feet intervals.
• The advantages of systematic random sampling are
  the ease with which samples are identified and
  collected and, sometimes, an increase in
  precision.
• The disadvantages of systematic random sampling
  are the poor accuracy and precision that can
  occur when unrecognized trends or cycles occur in
  the population.

65
Authoritative Sampling

• Authoritative Sampling - Sufficient information
  is available to accurately assess the chemical
  and physical properties of a waste, authoritative
  sampling (AKA judgment sampling) can be used to
  obtain valid samples.
• This type of sampling involves the selection of
  sample locations based on knowledge of waste
  distribution and waste properties (e.g.,
  homogeneous process streams). The rationale for
  the selection of sampling locations is critical
  and should be well documented.
• An example is an inspector taking one sample a
  discarded liquid that appears to be gasoline
  (color odor) to verify that it is gasoline and
  has a flash point below 140 F.

66
Enforcement Sampling RCRA Waste Sampling Draft
Technical Guidance, EPA530-D-02-002 (draft),
August 2002, page 10 11, RCRA online 50940

• 2.2.4 Enforcement Sampling and Analysis
• The sampling and analysis conducted by a waste
  handler during the normal course of operating a
  waste management operation might be quite
  different than the sampling and analysis
  conducted by an enforcement agency. The primary
  reason is that the data quality objectives (DQOs)
  of the enforcement agency often may be
  legitimately different from those of a waste
  handler. Consider an example to illustrate this
  potential difference in approach Many of RCRAs
  standards were developed as concentrations that
  should not be exceeded (or equaled) or as
  characteristics that should not be exhibited for
  the waste or environmental media to comply with
  the standard. In the case of such a standard, the
  waste handler and enforcement officials might
  have very different objectives.

67
Enforcement Sampling

• An enforcement official, when conducting a
  compliance sampling inspection to evaluate a
  waste handlers compliance with a do not exceed
  standard, take only one sample. Such a sample may
  be purposively selected based on professional
  judgment. This is because all the enforcement
  official needs to observe for example to
  determine that a waste is hazardous is a single
  exceedance of the standard.
• EPA530-D-02-002 (draft), August 2002, Page 11
  RCRA online 50940

68
Enough on Sampling?

• What about knowledge of process?

69
Quick Break

• Take 5
• Next Speaker John Misleh

70
Waste Determination
71
Waste Determinationby

• Process Knowledge
• Knowledge of Process
• (KOP)
• Generator Knowledge

72
Waste DeterminationCCR 66262.11

• (b) the generator may determine that the waste is
  not a hazardous waste by either
• (1) testing or
• (2) applying knowledge of the hazard
  characteristic of the waste in light of the
  materials or the processes used and the
  characteristics set forth in article 3 of chapter
  11 of this division.

73
Waste Determination

• CCR 66262.11
• Two options
• Process Knowledge
• Analysis

74
Knowledge of Process

• Why Use Knowledge
• Listed Waste is a function of how the waste is
  generated (knowledge)
• Know that it is Hazardous

75
Knowledge of Process OSWER 9938.4-03
RCRA Online 50010
76
Knowledge of Process OSWER 9938.4-03

• Process Knowledge -
• What goes in contaminants introduced what
  comes out.
• Waste Analysis Data from other facilities
• Old Analytical Data
• A lot of the information that is acceptable to
  demonstrate knowledge of process (KOP), looks a
  lot like Analytical Data

77
Knowledge of Process OSWER 9938.4-03

• Process Knowledge
• Material Balances
• Engineering Production Data
• Material Safety Data Sheets (1 10,000 ppm)
• Process Kinetic Information and Process Rates
• Other Engineering Calculations

78
Knowledge of Process OSWER 9938.4-03

• Analytical Data From Other Facilities
• Another plant that conducts the same process and
  is managing the same waste and has Analytical
  Data.
• A TSD that relies on waste analysis Analytical
  Data from offsite generators.

79
Knowledge of Process OSWER 9938.4-03

• Old Analytical Data
• Process and materials must be the same.
• Detection limits and equipment have improved.

80
Knowledge of Process OSWER 9938.4-03

• Situations where using KOP may be appropriate
• Constituents are well documented such as for F or
  K listed waste
• Wastes are discarded unused chemicals (P U
  listed)
• Health safety issues in sampling (too dangerous
  to sample)
• Physical nature of waste (construction debris)
  makes sampling impractical

81
Knowledge of Process OSWER 9938.4-03

• Conclusion (EPA Guidance Doc)
• Although EPA recognizes that sampling and
  analysis are not as economical or convenient as
  using acceptable knowledge, they do usually
  provide advantages. Because accurate waste
  identification is such a critical factor for
  demonstrating compliance with RCRA,
  misidentification can render your facility liable
  for enforcement actions.

82
Knowledge of Process Faxback 11918
Conservative Classification The regulations
allow a generator to characterize its waste based
on process knowledge, and it is understood that
generators may at times characterize their wastes
as hazardous conservatively, rather than incur
the costs of testing every batch or stream.
83
Knowledge of Process Faxback 11608
Analyzing Munitions not specifically required
to test their waste The determination may be
made by either applying knowledge of the waste,
the raw materials, and the process used in its
generation or by testing if they think that the
above munitions items would fail the TCLP-extract
analysis for lead or dinitrotoluene, then these
wastes could be declared as hazardous, and no
testing would be necessary.
84
Knowledge of Process Faxback 11592, 11579
Limited analytical Labs being unable to
determine conclusively that the waste is or is
not hazardous . . . It would probably be prudent
for the generator to manage those wastes as
hazardous waste.
85
Seeking Concurrence with DTSC? CCR 66260.200
(m) A person seeking Department concurrence with
a nonhazardous determination or approval to
classify and manage as nonhazardous a waste which
would otherwise be a non-RCRA hazardous waste
shall supply the following information to the
Department (5) laboratory results including
results from all tests required by chapter 11 of
this division and a listing of the waste's
constituents. Results shall include analyses from
a minimum of four representative samples as
specified in chapter 9 of "Test Methods for
Evaluating Solid Waste, Physical/Chemical
Methods," SW-846, 3rd Edition, U.S. Environmental
Protection Agency, 1986 (incorporated by
reference in section 66260.11 of this chapter)
86
Is ONE sample good for anything?

• Faxback 11907 - Representative sampling
  (Fluorescent Tubes)
• it appears that you tested one spent
  fluorescent tube to conclude that all of your
  spent fluorescent tubes are non hazardous. . . .
  Based on one tube, we have no way to assess the
  variability between fluorescent lamps. . . A
  representative selection of lamps randomly chosen
  should be analyzed to make this determination.

87
KOP Documentation OSWER 9938.4-03

• EPA looks for documentation that clearly
  demonstrates that the information relied upon is
  is sufficient to identify the waste accurately
  and completely.

88
KOP Documentation

• The generator is very familiar with the waste
  generation process and the California and Federal
  hazardous waste laws and regulations.
• Detailed chemical information for all the
  chemicals and materials utilized in the process
  is available.
• A detailed review of the generating process has
  been completed and the point of generation has
  been properly been identified.
• All documentation utilized to make the
  determination is included in the operating record
  associated with the waste stream.
• The generator has evaluated the information
  gathered and made a written determination.

89
Ten minute Break
Take a Break!
Please be back by