How to Calculate Daily and Weekly Noise Exposure (LEP,d and LEP,W) Using the HSE Method

Noise-induced hearing loss is the most-claimed occupational disease in UK construction history. Tens of thousands of civil claims have been brought against contractors over the past two decades, and the bill — paid mostly by insurers — runs into the hundreds of millions. The Control of Noise at Work Regulations 2005 (CoNAWR) require every employer to assess and control noise exposure, and the calculation method underpinning it all is, on the face of it, straightforward. The catch is that the regulations use a points-based shortcut that most SHEQ teams have heard of but few can derive from first principles, and a formal logarithmic formula that breaks half the Excel templates it touches.

This guide walks both methods end to end: the HSE points-based ready-reckoner (the fastest way to get a defensible number) and the formal logarithmic formula (for when you need an exact answer rather than a one-decibel band). It also covers weekly exposure (LEP,W), the often-forgotten peak sound pressure limit, and how hearing protection attenuation interacts with the Exposure Limit Value.

If you'd rather skip the maths, our free noise exposure calculator handles all of it — daily, weekly, points and logarithmic methods side by side. But understanding the method matters for defensible record-keeping, so this guide is worth the 15 minutes. This guide also sits inside our broader resource on UK occupational exposure calculations, which covers HAVS and whole-body vibration alongside noise.

What noise exposure measures

dB(A), LAeq and LEP,d — what the abbreviations mean

Noise exposure regulations use a small vocabulary of abbreviations that look intimidating but are straightforward once unpacked.

• dB(A) — A-weighted decibels. The A-weighting filter adjusts the raw sound pressure reading to match the frequency response of the human ear, which is most sensitive between roughly 500 Hz and 5 kHz. Almost every figure in a workplace noise assessment is A-weighted.
• LAeq — equivalent continuous A-weighted sound level. This is the steady noise level that would deliver the same total acoustic energy as the actual fluctuating noise over a given measurement period. It is what your sound level meter reports for a task.
• LEP,d — daily personal noise exposure. This is the worker's LAeq normalised to an eight-hour working day. A worker exposed to 95 dB(A) for two hours has the same LEP,d as a worker exposed to 89 dB(A) for eight hours.
• LEP,W — weekly personal noise exposure. The weekly equivalent of LEP,d, normalised over a five-day week. Used when daily exposure varies markedly day to day.
• LCpeak — peak sound pressure in dB(C). Measured with C-weighting (which preserves low frequencies that A-weighting filters out) and captures the instantaneous peak rather than a time-averaged level. Cartridge tools, explosive bolts and steel-on-steel impacts produce very high LCpeak values for fractions of a second.

The two figures the regulations care about are LEP,d (the daily eight-hour-normalised exposure) and LCpeak (the instantaneous peak). They are assessed independently — a worker can be under the LEP,d limit and still breach the peak limit, or vice versa.

The regulations

The Control of Noise at Work Regulations 2005 set three tiers of action, each with a corresponding LEP,d (time-averaged) and LCpeak (peak) figure.

• Lower Exposure Action Value (LEAV) — 80 dB(A) LEP,d or 135 dB(C) peak. At this level, the employer must provide information and training, make hearing protection available on request, and offer hearing health surveillance assessment.
• Upper Exposure Action Value (UEAV) — 85 dB(A) LEP,d or 137 dB(C) peak. At this level, the employer must designate Hearing Protection Zones, mandate the use of protection within them, implement a programme of noise reduction measures, and provide formal hearing health surveillance.
• Exposure Limit Value (ELV) — 87 dB(A) LEP,d or 140 dB(C) peak. The absolute ceiling. The crucial difference is that the ELV is assessed after the attenuation provided by hearing protection has been taken into account — it is the level at the worker's ear, not the level in the surrounding air. We cover this distinction in detail later in the guide.

These thresholds apply per worker per day. Group averages are not acceptable — assessments must reflect the exposure of identifiable individuals, even if those individuals are grouped into similar exposure categories.

The HSE ready-reckoner method (points-based)

The HSE publishes a points-based ready-reckoner that converts any combination of sound level and duration into a single number, with the points totalled across tasks to give the day's overall exposure. This is the fastest defensible method for routine assessments, and it is the method used inside the HSE's own Excel calculators.

How the ready-reckoner works

The system is anchored to one fixed reference: 100 exposure points equals an LEP,d of 85 dB(A) — the Upper Exposure Action Value. Every other figure on the table radiates outwards from this point using the standard noise-exposure energy maths.

• Lower EAV (80 dB(A)) corresponds to roughly 32 exposure points (more precisely, 31.62).
• Upper EAV (85 dB(A)) corresponds to 100 exposure points.
• ELV (87 dB(A)) corresponds to roughly 158 exposure points.

The advantage of the points system is that points are additive across tasks. dB values are not — you cannot add a 90 dB task to an 85 dB task and get 175 dB. Sound levels combine logarithmically, which makes adding them by hand cumbersome and error-prone. Points strip the logarithm out and leave you with simple arithmetic.

The HSE daily ready-reckoner (Crown copyright 2020) gives the points for any LAeq from 75 to 120 dB(A) at durations from 2 minutes to 12 hours. A handful of representative cells:

LAeq dB(A)	15 min	30 min	1 h	2 h	4 h	8 h
95	32	65	125	250	500	1000
90	10	20	40	80	160	315
85	3	6	13	26	50	100
80	1	2	4	8	16	32
75	0	1	1	3	5	10

The full table colour-codes results: green below the lower EAV, yellow between the lower and upper EAVs, red at or above the upper EAV.

Worked example — a typical site day

Take a groundworker on a typical UK site:

• Cut-off saw — 90 dB(A) for 2 hours
• Pneumatic nail gun — 97 dB(A) for 30 minutes
• General site exposure (banksman duties, plant in the background) — 82 dB(A) for 4 hours

Look up each task on the daily ready-reckoner:

• Cut-off saw, 90 dB(A) × 2 h → 80 points
• Nail gun, 97 dB(A) × 30 min → 100 points
• General site exposure, 82 dB(A) × 4 h → 26 points

Add them: 80 + 100 + 26 = 206 points.

Now look up 206 points on the right-hand side of the table. It falls between 200 points (88 dB(A)) and 250 points (89 dB(A)), much closer to 200, so the day's LEP,d is approximately 88 dB(A).

That is above the Upper EAV of 85 dB(A). The worker requires mandatory hearing protection, formal health surveillance, and a documented noise reduction programme. The exposure is below the ELV of 87 dB(A) only if attenuation from hearing protection brings it down — without protection, the worker is also above the ELV.

Weekly exposure (LEP,W) — when to use it

Some workers do not have a typical day. A site engineer might spend Monday on a demolition phase at 92 dB(A) and the remaining four days in the site office at 65 dB(A). Assessing each day individually paints a misleading picture of either danger or safety.

The regulations allow weekly averaging in limited circumstances:

• Where one or two days of the week are at least 5 dB higher than the others, or
• Where the working week comprises three or fewer days of noise exposure.

Outside those circumstances, daily assessment is the correct method.

The weekly calculation follows the same logic as the daily one, but over a five-day reference period. The HSE weekly ready-reckoner converts each daily LEP,d into daily exposure points (the same scale as the daily reckoner), sums them across the working days, and converts the total to an LEP,W figure normalised over five days.

Take a site engineer with this week:

• Monday — 90 dB(A) (demo cover) → 315 points
• Tuesday — 82 dB(A) → 50 points
• Wednesday — 84 dB(A) → 80 points
• Thursday — 81 dB(A) → 40 points
• Friday — 82 dB(A) → 50 points

Total: 535 weekly exposure points.

Look up 535 on the weekly side of the table: it sits between 500 points (85 dB(A)) and 630 points (86 dB(A)), closer to 500. The LEP,W is approximately 85 dB(A) — the worker has averaged exactly the Upper EAV across the week, even though four of the five days were below the Lower EAV. That demolition day pulled the weekly average up by itself.

This is the value of weekly averaging when used correctly: it gives a true picture of cumulative weekly exposure for workers with markedly uneven days. Used incorrectly — to smooth out a consistently noisy week and pretend it isn't — it is a breach of the regulations.

Run this in the noise calculator →

The formal logarithmic formula (for exact answers)

The ready-reckoner gives you a one-decibel band, which is fine for most routine assessments. The formal logarithmic formula is what sits underneath the reckoner and gives you the exact LEP,d figure to a fraction of a decibel.

When you need exact values rather than a band

There are three situations where the precise figure matters:

• Threshold cases. If the ready-reckoner result lands at 85 or 87 dB(A), you want to know whether the exact value is fractionally below or above the threshold, because the duties either side of it differ materially.
• Defensive documentation. For assessments likely to be scrutinised by an HSE inspector, an insurance adjuster, or a claimant's solicitor, an exact calculation is more defensible than a banded one.
• Trend monitoring. When you are tracking the effect of an engineering control over months — quieter compressor, isolated tool, redesigned workflow — you want to detect a 0.5 dB change. A banded reckoner reading hides it.

The LEP,d formula

The formal definition from Schedule 1 of SI 2005/1643 is:

For a single task of duration T at level LAeq:

LEP,d = LAeq + 10 × log₁₀(T / T₀)

Where T₀ is the eight-hour reference period.

For multiple tasks in a day, the levels combine on an energy basis, not arithmetically:

LEP,d = 10 × log₁₀ [ (1 / T₀) × Σ ( Tᵢ × 10^(LAeq,i / 10) ) ]

This is the formula that breaks Excel templates. It uses base-10 logarithms, exponents, and a sum across an arbitrary number of tasks — straightforward enough in code, but easy to misenter in a spreadsheet.

Run the groundworker example from earlier through the formula to verify the ready-reckoner result:

LEP,d = 10 × log₁₀ [ (1/8) × ( 2 × 10^9.0  +  0.5 × 10^9.7  +  4 × 10^8.2 ) ]
      = 10 × log₁₀ [ (1/8) × ( 2,000,000,000 + 2,505,936,000 + 633,957,000 ) ]
      = 10 × log₁₀ [ (1/8) × 5,139,893,000 ]
      = 10 × log₁₀ ( 642,486,000 )
      = 10 × 8.808
      = 88.08 dB(A)

The exact answer is 88.1 dB(A), which is consistent with the ready-reckoner's "approximately 88 dB(A)" result. Both methods agree — the ready-reckoner rounds to the nearest decibel, the formula gives you the decimal.

The same conversion works in reverse using the points system. Total points converted back to dB(A):

LEP,d = 85 + 10 × log₁₀(daily points / 100)

For our 206 points: 85 + 10 × log₁₀(2.06) = 85 + 3.14 = 88.1 dB(A). The same result.

Peak sound pressure (LCpeak) — the often-forgotten one

Daily exposure isn't the whole story. Time-averaged levels capture cumulative dose, but they do not capture brief, severe impulses that can damage hearing instantly. A single hammer-on-steel strike or a powder-actuated tool firing can deliver more acoustic energy in a millisecond than hours of background grinding.

The CoNAWR 2005 peak limits are separate from the LEP,d limits and are assessed independently.

• Lower EAV — 135 dB(C) peak
• Upper EAV — 137 dB(C) peak
• ELV — 140 dB(C) peak

LCpeak is measured in dB(C) — C-weighting, which preserves low-frequency content that A-weighting filters out. Peak values do not time-average; they capture the single highest instantaneous sound pressure during a measurement period. Even one event above the relevant threshold triggers the duties at that level.

Common construction sources that generate high LCpeak:

• Cartridge-operated and powder-actuated tools (fixing into concrete or steel)
• Explosive bolts and pyrotechnic actuators
• Steel-on-steel hammering (hammering masonry pins, knocking through formwork)
• Cut-off saw strike-throughs
• Compressed air bursts and pneumatic exhaust
• Demolition impacts on hard surfaces

A common mistake is to confuse LCpeak (peak C-weighted level) with LCeq (the equivalent continuous C-weighted level, used in the HML hearing-protection calculation). They are not the same and they do not interchange. LCpeak is typically 20 dB or more higher than LAeq for the same source; LCeq is usually only a few dB above LAeq.

Hearing protection and the ELV

Why the ELV is 87 dB(A), not 85

The Upper EAV (85 dB(A)) and the ELV (87 dB(A)) sit only 2 dB apart, which causes confusion. The key distinction is what level each refers to.

• The EAVs refer to noise in the air around the worker — the level a microphone would measure. They assume no hearing protection.
• The ELV refers to noise at the worker's eardrum — that is, the level after attenuation from any hearing protection is applied.

This means a worker can be exposed to 95 dB(A) of airborne noise but still be compliant with the ELV, provided their hearing protection provides at least 8 dB of real-world attenuation (95 − 8 = 87). Conversely, a worker on a 90 dB(A) site who wears poorly-fitting protection that only delivers 2 dB of attenuation in practice is above the ELV at the ear, even though the airborne level is below 91 dB(A).

The HSE convention is to apply a +4 dB "real-world" correction to any calculated protected level. This accounts for the gap between laboratory-measured attenuation (in a perfect, motivated, fully-trained user) and field-measured attenuation (in a tired worker who pushes plugs in halfway, or wears ear defenders over a hard hat that breaks the seal). If your manufacturer says a protector gives 25 dB SNR, treat it as delivering 21 dB in practice.

Three methods exist for calculating the attenuated level at the ear:

• SNR (Single Number Rating) — uses only the C-weighted workplace level and a single SNR figure from the protector's CE marking. Quickest, least accurate.
• HML (High–Medium–Low) — uses both the A-weighted and C-weighted workplace levels with three manufacturer-supplied attenuation figures. More accurate, especially when the noise spectrum is unusual.
• Octave-band — uses the unprotected noise spectrum in 1/1-octave bands and the protector's per-band attenuation. The most accurate method and the only reliable one for low-frequency-dominated noise (presses, generators, plant rooms).

The HSE strongly prefers HML or octave-band methods for any environment where the noise is not roughly typical of "general industrial."

For a deep walk-through of all three methods, with worked examples, see our companion guide on hearing protection performance assessment.

Where the manual / Excel method breaks down

The HSE ready-reckoner is a remarkable piece of design. It gives you a one-page lookup that produces a defensible LEP,d figure in two minutes. For one worker, one day, one site visit, it is genuinely faster than any software tool.

It is also where most UK SHEQ teams stop. And it is where the system starts to fall apart at scale.

• One worker, one day is the easy case. Most construction projects have 20 to 200 trades on site at any one time, each in a different exposure profile, often rotating between tasks day to day. Maintaining ready-reckoner calculations across a workforce is a full day's admin per week, and the calculations age fast as work packages change.
• Weekly exposure aggregation requires daily history. LEP,W needs each day's LEP,d before you can sum points. In an Excel-only world, that means a workbook per worker with a tab per week, and either a brittle web of cell references or a manual rekey every Monday morning. The first time someone copy-pastes over a formula, the data starts to drift.
• Peak exposure is poorly handled by templates designed for time-averaging. Most Excel noise templates ignore LCpeak entirely, or relegate it to a single column that nobody fills in. A worker who never breaches the LEP,d action values can still be exposed to a 142 dB(C) peak from a cartridge tool and the spreadsheet won't flag it.
• Hearing protection attenuation is its own calculation. If you are working to the ELV (rather than just the EAVs), you need to derive the protected level at the ear using HML or octave-band attenuation methods, apply the +4 dB real-world correction, then convert back to a protected LEP,d. Doing this in Excel is fragile — the formula uses both LAeq and LCeq, and the conditional logic between f ≥ 2 and f < 2 catches people out.
• Health surveillance needs longitudinal data. UK regulations expect employers to link exposure history to the audiometry programme. That means storing a worker's noise exposure record across months and years, alongside hearing test results. Excel does not do this well — files corrupt, version histories vanish, and the data set required for an HSE inspector or an insurance claim isn't available when asked for.
• Inspector audit trails. An HSE inspector wants to see who did the assessment, when, and what data they used. Excel cell-history isn't an audit trail. A signed, dated, locked record stored against an individual worker is.

How the RAMSGen noise calculator handles it

The RAMSGen noise tool was built specifically to address the gap between "I can do one ready-reckoner calculation in two minutes" and "I can manage noise exposure across a fleet of workers and projects without losing my mind."

• Both methods, side by side. Enter task LAeq and duration; the calculator returns points (ready-reckoner method) and the exact LEP,d (logarithmic method) in the same row. No need to choose between them.
• Daily and weekly aggregation in one place. Daily exposures roll up into weekly LEP,W automatically. Add a working day, the weekly figure updates.
• Peak exposure flagged separately. Tasks involving cartridge tools, impacts and other peak-heavy sources get a separate LCpeak input, with the 135 / 137 / 140 dB(C) limits flagged independently of the LEP,d result.
• Hearing protection attenuation built in. Toggle "HP worn" per task, enter the protector's H/M/L values once, and the calculator returns the protected level at the ear (with the +4 dB HSE real-world correction applied automatically) and the resulting protected LEP,d. Compares directly to the 87 dB(A) ELV.
• Per-worker history for health surveillance. Every calculation is stored against the worker, the date, and the assessor. The exposure record sits alongside the audiometry programme — the data set you need for an HSE inspection or an insurance audit is one click away.
• Audit-ready output. The calculator outputs a dated, locked exposure record you can attach to the worker's file or the project RAMS pack.

Open the noise calculator → — no sign-up required for a single calculation. For multi-worker tracking, weekly aggregation across a workforce, and audit-ready storage, see RAMSGen plans.

Noise FAQs

Is the ELV 85 or 87 dB(A)?

The ELV is 87 dB(A), but it is measured at the ear — after the attenuation from hearing protection is applied. 85 dB(A) is the Upper Exposure Action Value (UEAV), measured in the air around the worker, with no protection assumed. The two figures sound similar but mean very different things. A worker can be above the UEAV in the air and below the ELV at the ear if their hearing protection is working correctly.

Do I have to use the logarithmic formula, or is the ready-reckoner legally sufficient?

The ready-reckoner is legally sufficient for routine assessments. The regulations don't prescribe a method — they prescribe a result (a defensible figure for LEP,d). Most HSE-published guidance uses the ready-reckoner. The logarithmic formula is appropriate where you want an exact figure (threshold cases, trend monitoring, contested assessments), but it is not required for every assessment.

What about ambient site noise vs task noise?

Both count. The worker's LEP,d is the cumulative result of every period of noise exposure during the working day, whether they are operating a tool or standing 20 metres away while someone else is. Background site exposure at 80–82 dB(A) is often the largest single contributor to a groundworker's daily total, simply because they're in it for six hours. The ready-reckoner assumes you've broken the day down into discrete periods at distinct levels — what those periods are is up to your judgement.

How accurate are sound level apps on phones?

Useful for triage, not defensible for formal assessment. A calibrated phone app can tell you whether you're in a 75 dB or 95 dB environment, which is useful for spotting where you need a proper survey. It cannot give you a Class 1 or Class 2 measurement, it has no LCpeak capability worth speaking of, and the calibration is unverifiable. For an assessment that would survive HSE scrutiny, you need a calibrated sound level meter or noise dosimeter, ideally operated by a competent person.

When do I need a noise survey rather than estimates?

The regulations require a "suitable and sufficient" assessment. For most construction work, that means: a noise survey by a competent person whenever the work environment changes materially (new site, new plant, new process), supplemented by ongoing exposure calculations using survey data plus task durations. Estimates from manufacturer data and HSE conservative figures are acceptable as a starting point, especially during planning, but a formal survey is the gold standard for any prolonged or high-risk exposure profile.

Does weekly averaging let me hide a noisy week?

No. Weekly averaging is only valid where daily exposure varies markedly (one or two days at least 5 dB higher than the rest, or three or fewer days of noise exposure in the week). It exists to give a fair picture of cumulative exposure for workers whose days are genuinely uneven — not to smooth out a consistently noisy week. Used incorrectly it is a breach of the regulations and an inspector will spot it.

Further reading

• Pillar page: Occupational Exposure Calculations — UK Construction Guide
• Cluster page: How to Calculate HAVS Exposure — for workers using hand-held vibrating tools alongside their noisy ones (scabblers, breakers and cut-off saws are both noise and vibration risks)
• Cluster page: How to Calculate Whole Body Vibration (WBV) Exposure — for plant operators and drivers
• HSE source documents:
  • L108 — Controlling noise at work (Approved Code of Practice and guidance to the Control of Noise at Work Regulations 2005)
  • HSG258 — Controlling airborne contaminants at work (related where noise sources also produce dust)
  • INDG362 — Noise at work: guidance for employers (the plain-English summary)
  • HSE Daily and Weekly Ready-Reckoners (Crown copyright 2020) — the source tables underpinning this guide

Open the noise calculator →