1

Chapter 1 SAFETY ASSESSMENT

1. Introduction

Ordinarily it is difficult to quantify the assessment of safety conditions prevailing in

any organization. But such a quantitative evaluation can effectively provide a basis

for comparison and monitoring. In western countries, a few methodologies have been

formulated in this direction. In this chapter, we discuss some of those quantitative

methods, while applying these methods to Indian industries, it is important to gather

the relevant information particular to that industry.

1.1 Safety Performance Criteria

When appropriate performance criteria for safety are to be established, the two

important factors are the need for safety criterion and its associated necessary

properties.

Need: The various reasons why we would like to have safety performance criteria

are:

1. Carrying out comparisons

2. Estimating forecasts

3. Conducting trend analysis

4. Evaluating safety improvement program effectiveness

5. Identifying problem areas

6. Optimal allocation of resources for improving safety performance

Necessary properties: Two major properties are required in the performance criteria

are:

2

1. Reliability

2. Validity

Reliability: A criterion is said to be reliable if, for an error free measurement

technique, it is capable of duplication with the same results

obtained from successive applications to the same situation.

Validity: A criterion is said to be valid if, by its very nature, it is

satisfactorily suited to the object being measured.

1.2 Measurement of Safety Performance: Ingredients

For an accident to take place, the following must be present:

1. Worker

2. Machine, tools or equipment

3. Physical environment

4. Social environment

If AL denotes the “Accident Level” indicating the degree of losses, i.e., property,

injury etc. then:

( )( )( )( )�= gmeltkijL SPOWA

(Wij) = Combined effect of ith worker’s characteristics “j” as related to safety. Some

of the worker’s characteristics are physical, abilities and skills, interests,

and personality traits.

Physical characteristics: visual acuity, hearing, and so on

Abilities and skills: dexterity, verbal ability, intelligence, and mechanical

aptitude

3

Interests and Personality Traits: Scientific interests, safety attitude,

tolerance, and emotional stability

(Otk) = Combined effect of the tth machine’s characteristic “k” as related to safety.

Characteristics are

Mechanical Actions, Location of machine controls, quality of maintenance

(Pel) = Combined effect of physical environment “e” at location “l” as related to

safety. Some of the components are

temperature, humidity, illumination, air contamination, and the state of

housekeeping.

(Sgm) = Combined effect of social environment of characteristics “g” at location “m”

as related to safety. Examples of characteristics are,

Regulations, formal rules, laws that influence worker behaviour

1.3 Commonly Used Measures of Safety

The above equation provides a comprehensive analytical calculation of accident

level. There are some commonly used measurement techniques some of which are

presented below.

1.3.1 American National Standards Institute (ANSI) indices

The ANSI system uses frequency and severity rates that pertain to death, disabling

(lost time) injuries (including total, permanent partial, temporary total, and

temporary partial).

These measures pertain to the relative frequency of occurrence of major injuries and

days lost or charged to them.

4

The different indices of ANSI are:

1. Disabling Injury Frequency Rate (DIFR)

2. Disabling Injury Severity Rate (DISR)

3. Average Day Charged (ADC)

1.3.1.1 DIFR

DIFR is defined as the number of disabling injuries (including illness) per million

employee hours worked:

(# of disabling injuries) (one million) DIFR = ---------------------------------------------- (1.1)

(# of employee hours worked)

Example 1.1

Suppose in a given year, XYZ Company employed 500 fulltime workers and 200

half-time workers. The record of injuries and illnesses experienced by the workers

during that year are given in the table (E1.1)

Table (E1.1)

Type of injury or illness No. of injuries or illnesses Days lost or charged

Fractures 5 75

Ankle Twists 10 150

Thumb amputations at the

distal phalanx 4 1200

Cases of dermatitis 7 50

Total 26 1475

5

Calculate DIFR

Number of disabling injuries = 26

Number of employee hours worked = (Number of full time employees) x (40 hours /

week) x (50 weeks / year) + (Number of half time employees) x (40 hours / week) x

(25 weeks / year)

= 500 x 40 x 50 + 200 x 40 x 25 = 1,200,000 hours

Using the above calculations in equation (1.1) gives

(26)x (106) DIFR = ----------------- = 21.67

1,200,000

1.3.1.2 DISR

DISR is defined as the number of days lost or charged per million employee hours

worked.

(total days charged) x (106) DISR = --------------------------------------

(Number of employee hours)

Example 1.2: For the data of example 1.1., calculate DISR for the XYZ company

Solution: Number of days lost or charged = 1475

Number of employees hours worked = 1,200,000

Thus

(1475) (106) DSIR = ------------------ = 1229.16

1,200,000

6

1.3.1.3 ADC

ADC can be computed to indicate the average length of disability per disabling

injury as follows:

Number of days lost or charged DISR

ADC = ------------------------------------------- = --------- Number of disabling injuries DIFR

Example 1.3

Compute the value of ADC for the data of example 1.1.

1229.16 ADC = ------------- = 56.49

21.76

1.3.1.4 OSHA Rates

Occupational Safety and Health Administration (OSHA) introduced a new method of

measurement of safety. OSHA method of measuring is as follows.

Any occupational injuries and illnesses that result in death, regardless of time

between the injury and death or the length of illness, nonfatal occupational illnesses

other than fatalities that result in lost work days, and occupational injuries that result

in transfer to another job or require medical treatment are considerable recordable

cases.

OSHA suggests the flow chart shown in Fig. 1.1 for finding out recordable cases.

The incident rate is given by,

7

Number of recordable injuries x (200,000) IROSHA = -------------------------------------------------------------

Number of employee hours worked

It is to be noted in the above formula, the number 200,000 is the equivalent to 100

full-time employees at 40 hours per week for 50 weeks.

Example 1.4

A company employing 350 full-time employees experienced the accidents and

illnesses given in Table E1.4 during 1986. Calculate the incident rate.

(13) (200,000) IROSHA = ------------------------ = 3.71

(350) (40) (50)

Table (E1.4)

Injury / Illnesses Number Days Lost

Broken Ankle 1 25

Carbon Monoxide Poisoning 1 2

Dermatitis Cases 5 0

Welding Flash Burns 2 4

Cuts requiring stitches 4 0

Total 13 31

8

Fig. 1.1

1.3.1.5. Occupational Injury and Illness Frequency and Severity Rates

The illness frequency rate is defined as:

(Number of lost time injuries) (200,000) Illness Frequency Rate = -----------------------------------------------------

(Number of employee hours worked)

IF A CASE

Results From A Work Accident

Or From an Exposure in the

Does not result from work

accident or from an exposure in

A Death An Illness An injury, which involves

Medical Treatment

Loss of consciousness

Restriction of Work

Transfer to another job

None of This

Case must be recorded

The case is not to be recorded

9

The severity rate is defined as

(Number of days lost) (200,000)

Severity rate = ------------------------------------------------ (Number of employee hours worked)

Example 1.5

A company has two divisions, each employing 100 employees. In 1989, Division A

had 15 lost-time injuries and Division B had 3 deaths.

Calculate values of frequency rates.

Division A Davison B

Description Days

Lost

No. of Injuries

/illness

Description Days

Lost

No. of Injuries/

illness

1) Sprained Ankle 1 1 1) Broken Ankle 10 1

2.) Broken Thumb 2 1

3) Eye-injury due to embedded metal 0 1

4) Cuts requiring stitches 3 4

5) Loss of consciousness 0 2

6) Cases of dermatitis 4 4

Total 10 13 10 1

15 x 200,000 Frequency rate FR = ------------------- = 15 for division – I

200,000

3 x 200,000 Frequency rate FR = ----------------- = 3 for division – II

200,000

10

Example 1.6

For the company in example 1.5, Calculate the severity rates for both divisions

10 x 200,000 Division I : Severity rate = --------------------- = 10

200,000

10 x 200,000 Division II: Severity rate = -------------------- = 10

200,000

Apart from above, the following measures of safety are also being used by some

organizations:

1. Injury incidence rate

2. Illness incidence rate

3. Fatality incidence rate

4. Lost work days cases incidence rate

5. Number of lost work-days rate

6. Specific hazard incidence rate

In all these rates the standard 200,000 factor is used.

1.3.1.6 Limitations of these methods

There are some limitations to the existing methods of safety measurement.

1. The rates in question are not sensitive enough to serve as an accurate

indicator of safety effectiveness

2. The smaller the work force, the less reliable are the rates as an indicator of

safety performance

11

3. Lost time accidents, deaths and other recordable injuries are events of

relatively rare occurrence.

4. A single severe injury or death will drastically alter the severity rate in

smaller organizations, and this rate may not accurately reflect overall

accident prevention accomplishments

5. The rates do not reflect environments involving nonparallel hazard

categories.

6. The rates are based on after the fact appraisals of injury-producing accidents

1.4 Accident Cost Measurements

Accidents lead to losses. Some of these losses could be accident-investigation costs,

payments for settlement of law-suits, costs of corrective measures to stop

recurrences, payments for property damage on top of coverage by insurance,

insurance increment costs, legal costs associated with loss of public confidence (thus,

reduction in revenue). Awareness of these costs helps in justifying the associated

expenditures and manpower.

We discuss three of the many methods available.

1. Heinrich Method

2. Girmaldi and Simonds Method

3. Optimal Cost-Benefit Model

1.4.1 Heinrich Method

Heinrich’s Method states that normally the actual expenditure for the organization

due to an accident is five times the insured cost paid for that accident. This is

because of the presence of indirect costs.

The indirect cost elements enumerated by Heinrich are:

12

1. Cost of lost time of injured employees,

2. Cost of lost time of employees who stop work or are involved in the action

3. Cost of cost time by management

4. Cost of lost time on the case by first aid and hospital people not paid by

insurance

5. Cost due to machine / material damage

6. Cost due to lost orders,

7. Cost to employees under welfare and benefit system

8. Cost to employees in continuing wages of the insured

9. Cost due to weakened morale

10. Overhead cost for injured employee while in non-production status

Some or all of these costs may be relevant to a given incidence of accident.

1.4.2 Girmaldi and Simonds Method

In this method the total cost is divided into two categories; insurance costs and

uninsured costs. Insurance costs are not normally difficult to estimate. The uninsured

cost is given by,

Uninsured cost = A x (Number of lost work-day cases with days away from work

(lost days)) + B x (Number of doctor’s cases (OSHA non-lost workday cases that are

attended by a doctor) + C x (Number of first aid cases) + D x (Number of non-injury

accidents).

Here A, B, C, and D are constants representing the average costs for each case,

respectively.

Lost time cases include permanent partial disabilities and temporary total disabilities.

Doctor’s cases include temporary partial disabilities and medical treatment cases

13

requiring the attention of a physician First aid and those resulting in property damage

of less than Rs. 1000/- and in loss less than 8 hours of working time. No injury cases

include uninitiated occurrences resulting in loss of 8 or more man-hours or Rs.

1500/- or more property damage.

For finding out the values of the constants A, B, C, and D a pilot study may very

well be undertaken in an organization. Alternatively the Table (1,2) gives some

suggested values.

Table 1.2

Constant Suggested value (Rs.)

A 11,000

B 2750

C 600

D 20,000

The costs given in Table 1.2 were based on wage and price levels when average

hourly wage for production workers in manufacturing was Rs. 200/-. By dividing Rs.

200/- by the current average wage, a multiplier may be derived to adjust each of

these costs.

Example 1.7

The data given in Table (E1.7), were applicable to an organization. Calculate the

accident cost using these data and the Girmaldi and Simonds method.

The total accident cost (TAC) is given by

TAC = Insured Cost (IC) + Uninsured cost (UC)

14

The insured cost is defined by

IC = Insurance Premium – Insurance Refund

IC = 7500000 – 1350000 = 6150000

UC = A x 25 + B x 75 + C x 220 + D x 30

Thus, from the given information, the value of the wage adjustment multiplier

(WAM) is

1000 WAM = ------- = 5

200

The following adjusted values of A, B, C, and D are obtained:

A= 5 x 11000 = 55000

B = 5 x 2750 = 13750

C = 5 x 600 = 3000

D = 20000 x 5 = 100000

Using these values

UC = 55000 x 25 + 13750 x 75 + 3000 x 220 + 100000 x 30 = 6066250

The resulting total accident cost is

TAC = IC + UC = 6150000 + 6066250 = 12216250

1.4.3 Optimum Cost-Benefit Model

This approach aims at developing a cost-effective accident prevention program,

which maximizes the injury and illness reduction within constraints of available

15

funds. As more funds are expended for an accident prevention program such as

OSHA, the costs of injuries and illness should decrease. This relationship is shown in

Fig. 1.2.

The intersection of the two curves in Fig. 1.2 gives the minimum total cost when the

most benefit, in terms of reduced injury and illness costs, can be achieved for the

minimum investment.

Table (E1.7) Accident Cost-Related Data for an Organization

General Information Rupees Accident-related Occurrences Frequency

Average Production Worker

Wage (Hourly) 20 - -

Insurance Premium 150000 - -

Insurance Required 27000 - -

- Lost time cases 25

- Doctor’s cases 75

- First aid cases 220

No injury cases 30

1.5 Statistical Analysis of Safety

Safety professionals’ ability to measure, predict and control potential factors for

minimizing losses in an organization is enhanced by performing statistical analysis

of the raw safety-related data.

16

1.5.1 Safety Behaviour Sampling

Safety behaviour sampling is a technique of measuring unsafe acts.

Total Cost Protection/prevention cost Cost Injury / Illness cost High Low

Number of Injuries or Illnesses Figure 1.2

There can be two causes of errors:

1. Error-committing characteristics of people

2. Error-provocation situations.

By providing necessary feedback to people concerning their errors, they can be

enabled to reduce their error-committing characteristics. Fig. 1.3 shows the influence

of various factors on human performance.

Unsafe acts are errors made by workers, and unsafe conditions are existence of error

provocation situations. According to Heinrich, roughly 88% of all accidents are

made by unsafe acts of people, 10% are made by unsafe conditions and 2% by the

acts of God. A study conducted on a total number of accidents of 551 has revealed

many crucial facts. The classification of causes of accidents shows that 30% of these

accidents occurred due to poor maintenance. That is 167 accidents. Next major cause

17

was found out to be mis-operation leading to 29% or 162 accidents. Accidents

resulting from poor supervision were 14% or 77, resulting from unsafe acts by

worker were 6% or 33, resulting from incomplete check were 5% or 30, resulting

from inferior design or defective fabrication were 5% or 26, and rest were due to

reasons like outside disturbance and others and ultimately due to the acts of God or

due to unknown causes. The later reasons contributed to no more than 6% or 32

accidents whereas the rest all accidents of 94% or 519 could be attributed to some

cause. In another classification, it can be concluded that 55% or 303 accidents were

due to Human factors (mis-operation, poor supervision, unsafe acts and incomplete

check) and 39% or 216 accidents were due to Physical factors (poor maintenance and

inferior design or defective fabrication) and 6% or 32 accidents were due to other

causes (outside disturbance and acts of God).

This study shows results that fall near to the Heinrich’s classification with a small

deviation.

1.5.1.1 Fundamentals

Safety behaviour sampling is based on the laws of probability. If we are dealing with

a process that can be only in two states, safe and unsafe, the total probability is 1 or

100%. In a multi-activity study each observation is in a binary state for each activity

considered.

In terms of probabilities, we can express the relationship as

p + q = 1

p = the probability of a single observation in one state, say S for safe act

q = (1 - p) = probability of no observations in state S

For “n” observations

18

(p + q)n = 1

where n is the number of observations in the sample.

The distribution of probabilities resulting from the binomial expansion follows the

binomial distribution. The mean of the distribution is “np” and the standard deviation

of the distribution is )p1(np − . As “n” becomes large the distribution becomes

almost continuous and takes as the properties of normal distribution. When “n” is

large and neither “p” are “q” are close to zero, the mean and standard deviation are

obtained by:

Sample mean = np/n = p; sample standard deviation = n)p1(np −

n)p1(p −=

This sampling technique has demonstrable usefulness in evaluating unsafe

behaviour.

Here, it is assumed that the percent of time a worker working safely / behaving

safely, can be determined.

In order to obtain a complete and accurate picture of safe / unsafe acts performed by

the worker, it is necessary to continuously observe the worker and record data related

to unsafe acts. Note that a sufficiently large sample must be obtained for

representative results.

For a large number of observations, the resulting distribution approaches the shape of

a normal curve.

19

You have learnt about quality control in last semester. You may recall that, 95.26%

of all observations fall within 2 standard deviation limits and 99.74% within 3

standard deviation limits. It is emphasized that the normal distribution may only be

assumed for a large number of observations. One must determine the adequacy of the

number of observations in terms of the desired level of confidence. Generally, a

confidence level of 95.26% or within 2 SD, is considered adequate for most safety

behaviour sampling studies.

This confidence level means that the conclusions will be representative of the true

population 95% of the time.

In addition to the confidence level, the safety behaviour sampling has another

attribute called accuracy. Accuracy may be interpreted as the tolerance limit of the

observations that fall within a desired confidence level. 5% accuracy with 95%

confidence level is the combination often used in safety behaviour sampling. This

means that 95% of the time within 5% accuracy limit, the conclusion drawn on the

basis of safety behaviour sampling will be representative of the actual population.

Safety behaviour sampling requires randomness of occurrences. This is achieved

when each period of the workday is equally likely to be selected as the observation

period.

1.5.1.2 Procedure for Safety Behaviour Sampling

1.5.1.2.1 Define Work Stations

This includes departments / units in an organization where safety behaviour sampling

is to be conducted: hot room, print press section, stock and shipping.

20

1.5.1.2.2. Prepare a List of Unsafe Acts

This list can be developed from plant accidents recorded initially and modified later

as appropriate. Plant accidents include all accidents, such as disabling injuries,

recordable injuries, and first aid cases. A specimen worksheet is given in Fig. 1.4.

Fig. 1.3

Environment Skill, Training, Motivation, Physical condition

Human Performance Errors in Human

Performance

Worker Capability

Performance Evaluation Standards

21

Fig. 1.4

1.5.1.2.3 Conduct a Pilot Study

Prior to conducting a pilot study, one must carefully select times to observe worker

behaviour. These times must be selected randomly. The number trial observation

periods required depends upon the number of persons observed. For a guideline

purpose, it is said that a sufficient number of trial observation periods should be

selected so that the total sample size is at least 100.

When planning trial observation periods, there are many methods available for

arranging observation periods in a random pattern. One such method is described

below.

SAFETY BEHAVIOUR SAMPLING WORKSHEET

Department: Type of Activity: Work Center: UNSAFE ACTS 1. Handling Hot Parts with Unprotected

Hands 2. Failure to wear proper safety glass 3. Improper Lifting 4. Carrying heavy load --------------------- --------------------- --------------------- Total Unsafe Acts

Date: Time: Safety Analyst Signature:

22

In this method, the first hour of the working day is identified by the numeral 1 and

the second by the numeral 2 and so on. A table of random numbers is used to obtain

a series of three digit figures, the first digit representing the hour of the working day

and the next two the minutes. Numbers representing hours not in the working day, or

impossible minute values, are discarded. A sufficient quality is obtained to give the

required observation times for each day of the study. There should always be a

separate list for each day.

Example 1.8

Assume that a plant operates from 8 am to 5 pm. In this case, the time 8 am is

designated by the first digit of the random number, 9 am by 2, and so forth, down to

4 pm by 9. A short random number and its interpretation is given table (E1.8).

Table (E1.8) A Random Number Table and Interpretation

Random No. Interpretation

907 4.07

882 Impossible minute value – discard

544 12.44 pm

720 2.20 pm

838 3.38 pm

010 Impossible hour – discard

413 11.13 pm

For actual use, the list of observation times is to be arranged in time sequence.

1.5.1.2.4 Observer Training

The observer must be instructed to categorize worker behaviour as being either safe

or unsafe as defined by the entries of behaviors included in the unsafe acts list. He /

23

She should make a trail run and practice how to decide instantaneously whether the

observed behaviour is safe or unsafe. In addition, the observer should be trained to

determine whether the behaviour of each worker is safe or unsafe at the time of each

observation. While observing a department, the observer should walkthrough the

department and see whether workers are behaving unsafely.

1.5.1.2.5 Calculation of Required Number of Operations

The number of observations required is based on data collected during the pilot

study, the degree of accuracy required, and the given level of confidence.

Two terms are recorded during the pilot study:

1. Total number of observations made (N1)

2. Number of observations in which unsafe behaviour was observed (N2)

Thus, the proportion of unsafe behaviour is,

pNN

1

2 =

If S = desired accuracy

N = Total number of observations required

K= the value obtained from standardized normal tables for a given level of

confidence, then

( ) ( )p1pSKN

2−=

For a given level of confidence K, the value of K is read from the standardized

normal tables. For 95% confidence, K is approximated as 2, and for 99% confidence,

K is taken as 3.

24

Thus, for a 95% confidence,

( ) ( )p1p54N

2−=

Example 1.9

Assume that a pilot study generated the following results:

Total observations = 120 = N1

Unsafe observations = 35 = N2

Thus, we have

029.012035p ==

Now, assume that we want to calculate N for a 10% accuracy and 95% confidence

level. Thus substituting the above given values,

( ) ( ) 8229.0129.01.02N

2=−××=

This means that a minimum of 82 observations are required to produce satisfactory

results.

1.5.1.3 Conducting Additional Observations

Carry out the actual study by making (N-N2) number of random observations and

recording the behaviour according to safe and unsafe classifications.

25

After conducting the additional observations, calculate S by using the following

formula:

( ) NppS −= 1

In the case of the above example, we have K = 2 and N=82.

If 1.0S ≤ , the results obtained provide us the desired accuracy and confidence level;

in case S > 0.1, calculate new N and start making the observations as explained

above.

This study should be repeated once a week for a series of 6 to 12 weeks. Number of

repetitions depends upon the sample size N, the number of persons observed, and the

available manpower to undertake the study.

1.5.2 Correlated Work

When several workers are observed at (or near) the same time and their observations

are summed into the same x values (and, hence, p values), then their individual

reading can not be considered to be independent. Even if the workers appear to act

independently, their activity is correlated (not independent) within an observation

round, simply because the observations occur at the same time of the day. For

example, an observation round at the end of the day is likely to find all workers in a

cleanup mode.

This correlation can be compensated for by calculating for the Kth category by

alternate standard deviation

( ) BB VSD =

26

where

( )[ ] ( )

( )1JN

kNpjmj,kYV

J

1j

22

B −

−=�

=

J = total number of observation rounds made during a study

M(j) = total number of workers (or machines) observed on the jth observation round

of the study. [For example, m(5) is the number of workers observed on the

fifth observation round of the study. Ideally, this quantity should be

constant throughout a study, but, in practice, it is rarely the case).

Y(k,j) = total number of workers (or machines) found in the kth work category on the

job observation round. [For example, Y(3,15) is the number of workers

found in category 3 on the 15th observation round of a study. If only 1

worker is being observed, Y will only be “0” or “1” for each k and j].

Symbols N and p were defined earlier.

Example 1.10

Calculate (SD)B for the data given in Table (E2.3).

Solution:

When, as in the example, only 16 observation rounds are made in a total study, it is

appropriate to compensate for the small number by using a percentage point from “t”

distribution instead of “z” distribution. In practice, however, more than 30 rounds of

a study are to be made, and the difference between a “t” value and a “z” value is

trivial.

27

Thus, from the table (E2.3)

P1 = 41 / 232 = 0.177

N = 232

And

J = 16

Thus

( )[ ] ( )[ ] 001709.0116232177.0232217.13V 2B =−×−=

Thus

( ) 0413.0VSD BB ==

Table (E1.10) given Data values and Relevant Calculations

Round No. of Workers Observed m(j)

No. of Workers observed Y(i,j) Y(i,j)2/m(j)

1 10 7 4.900 2 10 4 1.600 3 14 5 1.786 4 14 2 0.286 5 14 3 0.643 6 16 0 0.000 7 16 4 1.000 8 16 1 0.063 9 16 1 0.063

10 16 2 0.250 11 16 3 0.563 12 16 4 1.000 13 16 0 0.000 14 16 1 0.603 15 16 4 1.000 16 10 0 0.000

Total 232 41 13.217

28

1.5.3 Safety Behaviour Control Chart

For developing a control chart, the mean of p (fraction of time each worker is

involved in unsafe acts or the mean percent unsafe behaviour of the entire group

during the observation period) is computed, using p, the upper control limit (UCL)

and the lower control Limit (LCL) can be computed with the aid of the following

expressions:

( )( ) Np1p2pLCL

Np1p2pUCL

−−=

−+=

where p is the mean of the p’s.

The ‘2’ appears because a 95.44% confidence level is to be provided.

Example 1.11

Table (E2.4) shows the results of a sampling study for 50 observations per day. The

table shows proportions of unsafe acts observed each day. Using the data o the first

10 days, obtain the values of UCL and LCL o a safety behaviour control chart.

Table (E1.11) p Values for Samples of 50 Observations Each

Proportion of Unsafe Acts

Day p Day p Day p 1 0.16 6 0.16 11 0.15 2 0.14 7 0.18 12 0.16 3 0.13 8 0.15 13 0.18 4 0.17 9 0.15 14 0.17 5 0.17 10 0.14 15 0.45

29

Solution:

The mean p value for the first 10 days = 0.155

Using the above resulting value and N = 50

( )

( )055.010.0155.0

50155.01155.02155.0LCL

257.010.0155.0

50155.01155.02155.0UCL

=−=−×−=

=+=−×+=

This shows that the values for p as given in Table (E2.4) for the first 10 days are well

within limits of UCL and LCL.

1.5.4 Improving Safety Behaviour

In order to improve safety behaviour of workers, a major program must be

introduced. This could be comprised of safety training programs, lecture series etc.

The safety behaviour sampling study may be conducted on a weekly basis during and

after the completion of the program. The safety behaviour control chart for each

period following the start of the program will show if a significant improvement in

unsafe behaviour has been achieved. Modification of the program or components o

the program may be carried out as long as the unsafe behaviour is being reduced.

Once the minimum of unsafe behaviour has been achieved (i.e. p), the behaviour

sampling study may be repeated and the data plotted on the control chart to assure

that unsafe behaviour remains at the desired minimum level.

30

Chapter 2 WORK ENVIRONMENT

1. Introduction

Ensuring that the employees work in comfortable work environment is an essential

part of the safety management in industry.

The following conditions can be termed as poor work environment.

• Excessive noise levels

• Poor lighting

• Exposure to heat stress

• Poor ventilation

• Presence of toxic or irritant contaminants in the atmosphere.

2. Noise

Noise can be described as an unwanted sound. It is difficult to precisely separate

noise from pleasant sound. A 3500 watt rock music at full volume may be noise for

an elderly person but the same may not be the noise for a teenager. The ear of a

normal person can respond to sound waves in the frequency range of 20 cycles/

second to 20000 cycles / second. The sensitivity is greater in the range of 2000 c/s to

5000 c/s.

2.1 Effects of Noise on Human-beings:

Following are the main ill effects of noise on persons:

31

• Loss of hearing capacity: High noise levels like that of an explosion can

lead to complete hearing loss instantaneously. Exposure to noise levels above

safe limit can lead to gradual hearing loss over a few years. Recommended

limits for noise exposure levels can be ascertained from national/international

standards on it. At person exposed to a noise level of 100 dBs daily for 8

hours will lead to hearing loss over a few years.

• Reduced work performance: In noisy work environment, the chances of

errors in general and particularly work involving mental concentration and

skills are high. Further, the communication becomes difficult and this can

pose serious hazards.

• Physiological effects: Some of the ill-affects of high noise levels on body

functions are: rise in blood pressure, increased heart beat, muscular and

nervous tension, loss of sleep, indigestion and increased fatigue.

• Psychological effects: Annoyance and lack of concentration are predominant

effects of high noise. But these effects by and large, have a bearing on the

predisposition of the individuals too. e.g. an operator who gets a much

desired promotion might consider the noise of his machine as music whereas

the hissing sound of leaking steam might be highly annoying to a

discontented worker. However, it should be remembered, severe

psychological disorders can result from exposure to high noise levels.

2.2 Unit of sound

The rate of flow of sound energy or the intensity of sound can be measured in

watts/ meter2 which is the fundamental unit of sound. However this is not a

practical unit. A practical unit (BEL) is derived by comparing the sound intensity

with the lowest intensity which the human ears can sense and then using a

logarithmic scale. This can be expressed as

32

Intensity level in BELS ���

����

�=

heardbecanthatintensitysoundlowestw/minintensitySound

log2

10

���

����

�= −12

2

10 10/intensity

logmwinSound

A modified unit of sound called decibel is introduced to make the noise level

within 0-200. One decibel is equal to 10 dBs.

Intensity level in decibels = 10 ���

����

�−12

2

10 10/intensity

logmwinSound

This is the commonly used unit of sound and is abbreviated as ‘dB’.

Some typical Noise Levels

Background noise in TV Studio 20dB

Residential area at might 40dB

General offices 50dB

Normal conversation 60dB

Heavy lorry accelerating 90dB

Printing Press Room 100dB

Sheet Metal Shop 100-120 dB

Compressor Room 110-120dB

Turbine Room 110-120 dB

Engine Room 100-110dB

Fitting and chipping 100-110dB

Sound can be measured by direct reading instruments. The most simple among

sound measuring instruments is: Sound level Meter. This meter has a

33

microphone, which transforms sound pressure variation into electric voltage

fluctuations and an amplifier to enlarge the voltage fluctuations sufficiently to

activate a meter. The meter is calibrated to read the sound intensity in decibels

directly. However, this meter cannot split up the sound levels into different

frequencies. It simply adds up all the sound levels in different frequencies. But,

for diagnosis of noise problem and for selecting efficient control measures, the

frequency spectrum of noise levels is required. For this purpose an Octave Band

Analyzer is used in conjunction with the sound level meter. With an Octave Band

Analyzer, a particular frequency range can be selected and the sound level in that

frequency alone can be measured.

There are many industrial situations where a person is not exposed to same noise

level during working hours such as, in power plants an employee is required to

work in different areas with different noise levels (control room,

turbine/compressor area). In such situation, equivalent continuous sound level

(ECSL) would be useful from the point of hearing protection. ECSL is the

continuous sound level that in 8 hours (normal shift time) of exposure would

produce the same effect as the exposure to varying levels during this time. There

are integrating meters, which can automatically compute the ECSL. An operator

can carry these meters with him during working hours. These meters are also

called “dose” meters since these measure the dose which the operator is exposed

to in a day.

2.3 Control of Noise

Almost every industry has some processes or machines, which produce noise.

Many factors affect the techniques or mechanisms that can be used to control

noise. First, one has to find out the level of noise, the frequencies involved and

the desired reduction in noise level. Then one has to find out how the noise is

generated, how it spreads, the path ways through which noise is transmitted and

the surfaces which can reflect noise. Some of the noise control methods are:

34

Reduction at Source

Wherever possible, it would be advisable to reduce the noise at the source

itself. There are many situations where a sizeable reduction in noise level can

be achieved by this method. Some examples are:

Bearing making noise due to wear – change bearing.

Excessive load on bearing – change the size of bearing

Leaks in compressed air line steam pipes etc – stopping the leak.

Exhaust of diesel engines – install silencers

Vibration of heavy rotors – Dynamic balancing of rotors and better

design of foundation.

Vibration isolation

In heavy machines, mechanical vibrations are transmitted through the

structure, walls or the floor. This results in vibration of other parts of

machines, fittings (conduct pipes, ventilation ducts) and sheet metal

components of equipments. This increases the overall noise level in the work

environment. In such situation it is necessary to stop transmission of

vibration from machine to floor and structure to reduce the noise. This can be

done by:

• Using vibration resilient mounts to fix the machine to foundations.

• Special heavy foundations with a large weight compared to the weight of

machine.

• Installing machines (particularly reciprocating type) a few feet below the

floor level of the work piece.

35

It is also important to isolate any direct solid link such as electric cable

conducts, water lines, steam pipes etc as vibration gets also transmitted

through these links.

Vibration Damping

When the frequency of vibration coincide with the natural frequency of

vibration, a phenomenon called, resonance takes place. Machine parts,

ventilation ducts and parts of structure cause noise in this manner. The noise

in these cases can be substantially reduced by damping – stiffening the

member or by applying a layer of material with a high hysteresis loss to one

side e.g. the ringing of long pipes can be eliminated by covering these first

with a layer of mineral wool and then with a hard casing of cement.

Similarly, sheets (MS) can be stiffened by M.S. angles. Noise from sheet

metal doors, sheet metal parts of machines and partitions etc. can be reduced

by this method.

Silencers

Silencers can be used effectively to reduce noise due to movement of gases or

air. e.g. exhaust silencers of automobile engines. Similarly silencers can be

used to reduce noise in the compressor inlet/outlets,turbine outlets (open),

exhausts, release of steam and gases and pressure relief values of pneumatic

machines. Silencers are comparatively cheap.

Noise Insulation

Sound waves also travel through air, therefore, substantial reduction in noise

can be achieved by insulating the source from the outside environment or in

other words erecting a barrier made of noise insulating material in the path of

noise so that the noise waves are reflected back. Insulation from all sides is

36

necessary for best results, however, insulating from two or three sides also

give good results. Closing from all sides often involves problems relating to

heat dissipation, accessibility; air, water and electric connections. The

reduction in noise depends upon the reflecting properties of insulating

material, the thickness of insulations, gaps or other path ways left open in the

barrier.

Noise absorption

Noise absorption materials prevent reflection of noise and also convert some

of the noise energy into heat energy. You must have seen the walls & roofs of

cinema halls and lecture theatres covered by some porous and soft materials.

This is done to absorb noise. A combination of absorption and insulation

often can give very good reduction in noise levels.

It should be remembered that no single method of noise control might give

satisfactory noise reduction. In many cases more than one of the above methods

discussed have to be used to achieve required reduction in noise.

Points to be remembered during planning stage of the plant for

noise consideration

• Procure equipments & machinery with low noise levels.

• Plan the layout such that the equipments with higher noise

levels are separately housed

• Design foundations, ventilation ducts, service connections

(air, water, electricity) taking into consideration the noise

control measures proposed.

37

If noise considerations are not thought of during planning stage then changes may be

necessary after installation of machines and equipments, which works out to be very

expensive. In some cases, noise reduction can be brought about by simple and

inexpensive methods. However, this is not the case always and often noise control

measures are limited by layout of machines and buildings, considerations of

ventilation and heat dissipation, accessibility for operations, inspection and

maintenance problems, and of course cost considerations.

2.4 Protection against Noise

When noise control measures are not practical, protection against noise has to be

considered. Following are few viable methods:

• Reduce exposure time: Timings should be so adjusted that workers in a

particular group work for a short time in high noise level areas and then

go to a quieter area. The objective is to keep the ECSL within safe levels.

• Vibration isolation: Isolation of machine/equipments that produce noise

is a good alternative e.g. in the operation of compressors and turbines, the

operators are not required to be near the machine all the time and have to

go near the machines only periodically to inspect and take readings.

Providing a separate insulated room for the operators to stay and watch

the machines for the remaining period can be a practical and cheaper

solution as compared to control measures for reducing the noise itself.

• Use of personal protective instruments: When the period of exposure

cannot be controlled, consider the use of personal protective instruments:

ear plugs and ear muffs. Use of ear muffs and plugs is recommended in

all industrial operations involving high noise levels like compressor

rooms, generating rooms, engine rooms, sheet metal shop etc.

An ear plug (soft button like piece made of wax impregnated cotton wool

or of soft plastic or of soft spongy material) when inserted into the ear fits

38

snugly into the ear canal and a noise reduction of 2 to 12 dB can be

achieved. Proper fitting of the plug is important. A loose fit will not give

any reduction and the tight fit will make it uncomfortable to wear.

The ear muff consists of two rigid cups or shell(The hard shell outside

and the soft acoustic material inside ) joined together with a head band..

Muffs can give noise reduction of 10 to 15 dB.

3. Heat Stress

In many industries, the employees are exposed to high or low temperatures e.g.

working near furnaces, boilers, chilling plants etc. Both the situations cause

discomfort and can lead to health impairment and affect performance. Proper safety

measures should be taken to safeguard the employees from heat stress. Generally,

over exposure to heat produce more ill effects than exposure to cold. So, only the

effects of heat will be discussed.

3.1 Effects of heat exposure on Human-beings

When one is exposed to heat, physiological changes takes place automatically

prompting the person exposed to take corrective action. But in case exposure time is

too long or the conditions so severe that the regulatory mechanism within the body

cannot cope up with the changes, temperature of the inside organs will rise that is

dangerous. Following ill-effects on the human body are caused by the exposure to

hot environment.

• High blood pressure. When the body has to cope up with high temperature it

necessitates increased blood circulation thereby increasing work load on

heart.

39

• Loss of water and vital salts from the body. At higher temperatures more

sweat is secreted from the body thereby depriving the body of vital salts and

water. This can cause exhaustion and impairment of body functions.

• Physiological disorders like indigestion. This is because on one hand blood

plays an important role in the dissipation of heat and heat regulating process

and on the other hand blood is related to most of the body functions.

• Physiological changes take place to prevent muscles from producing heat.

These changes are associated with tiredness, tendency to sleep and fatigue.

The person also ceases to be alert and his capacity for work gets deteriorated.

This lead to errors and accidents.

Which is more uncomfortable? A May day with 44°C temperature or a July monsoon

day with 37°C temperature in Delhi. Everybody is going to say the second options is

more uncomfortable even though the temperature is lower. Why? It is so because

during monsoon the humidity is more and also air velocity is almost negligible. It is

clear from this example that not only temperature but also humidity and air velocity

are also parameters of heat. Effective temperature is a unit which takes into account

the temperature, humidity and air velocity and is directly read from ready-made

charts using reading of Globe thermometer (radiant heat), dry bulb thermometer (air

temperature), wet bulb thermometer (humidity) and air velocity. WBGT is an index

that takes into account all the three factors. WBGT index is calculated as follow:

WBGT = 0.7WB + 0.2 GT + 0.1 DB

Where

WBGT is WBGT index

WB is wet bulb thermometer reading

GT is globe thermometer reading

DB is dry bulb thermometer reading

A WBGT index of 79°F is considered a comfortable level for most people.

40

Do you know the various types of thermometers and their uses. If yes, well, if no,

read from physics books of secondary (or matric) class or ask your instructor during

his visit.

3.2 Combating Heat

Before one take measures to combat heat, one has to identify the source of heat and

actual temperature conditions. This should be followed by assessing the effects of

heat on persons exposed to get an idea of extent and magnitude of the problem.

Following are few measures to combat heat:

• Insulation of hot surfaces like steam pipes can stop the atmospheric air from

getting heated.

• Separation of heat sources from general working area can prevent exposure to

many persons e.g. a furnace in a shop can be shifted to a small room or

enclosed in brick walls from the four sides to prevent exposure of persons

other than the operator.

• Radiation shielding can be applied if the heat stress is due to radiant heat.

Generally, a sheet of reflective material is placed between the source of

radiant heat and the worker(s). This principle can be applied in general to

decrease the temperature inside buildings by having shiny, reflective roofs

and/or southwest walls.

• Adjusting humidity can improve the situation in dry climates. During, humid

and hot conditions providing air velocity can bring some comfort.

• If improvement of conditions is not possible by any of the above methods

then one has to think of the personal protection like limiting exposure time

(as discussed in noise control) by allocating work in such a manner that

persons after working for sometime in hotter areas will be moved to colder

areas. Introducing rest pauses in between work at suitable intervals, proper

41

salt and water consumption and heat reflecting aprons are some personal

protection methods thought of

4. Industrial ventilation and exhaust systems

4.1 Industrial ventilation

In industries many persons working together causes air pollution (human body

consumes oxygen and releases carbon dioxide, sweat and body odour contamination,

human body dissipates heat) Apart from this, machines and processes also release

heat. Dust, gases and fumes are also generally generated in the process which make

the work environment uncomfortable and unsafe.

Ventilation improves the quality of air around the worker by replacing the

contaminated/ foul air by fresh air from outside. Broadly there are two types of

ventilation

• Natural general ventilation

• Mechanical general ventilation

Natural General Ventilation

It means the replacement of inside air by natural air without the aid of any other

external equipment. During design stage itself enough air inlet and outlets are

provided at strategic locations so that natural wind can pass through the area

automatically. Various factors to be considered during plant design for efficient

natural ventilation are: position of building with respect to wind direction prevailing,

position of sun and admission of sunlight, outside temperature during different parts

of the year and the desired inside temperature conditions. Special types of ventilators

that increase the natural flow of air are also commercially available and can be used

ERROR: syntaxerrorOFFENDING COMMAND: %ztokenexec_continue

STACK:

-filestream-