# GAC Bed Design for Drinking Water: Media Depth & EBCT

> Source: https://www.yelicarbon.com/blog-gac-water-design.html

Most GAC problems in drinking-water plants are not carbon-quality problems — they are bed-design problems. Wrong contact time, shallow beds, or no plan for media decay turn good carbon into disappointing results.

[GAC contactor bed design parameters for drinking water treatment]

**By the YELI Technical Team** · Updated August 2026 · 8 min read

**Short answer:** A drinking-water GAC contactor is sized on three numbers: empty bed contact time (EBCT) of 7.5–15 minutes for taste-and-odour and moderate organics (15–20+ for PFAS), media depth of 1.5–3 m, and hydraulic loading of 5–15 m/h. Common carbon specs are 8x30 or 12x40 mesh, iodine 900–1100 mg/g, hardness ≥98% (coconut), moisture ≤5%. Replace media on data — iodine decay, breakthrough before design EBCT, or rising pressure drop — not on a calendar.

Every operator buying granular activated carbon (GAC) for drinking water gets a quote full of carbon specs — mesh, iodine, hardness. What often goes missing is the design context: how deep should the bed be, how long should water stay in it, and when should you replace the media? This guide gives the numbers to check before you sign, and the numbers to monitor after start-up.

## The Three Numbers That Matter

Three parameters control most of the performance of a GAC contactor, and you can specify them without any software:

- **Empty bed contact time (EBCT)** — the volume of carbon divided by the flow rate, in minutes. For taste-and-odour compounds (geosmin/MIB) and moderate organics, plan for **7.5–15 minutes**. Treat 5 minutes as the absolute floor; for tougher duty such as PFAS, designs often use 15–20 minutes or more.

- **Media depth** —**1.5–3 m** is the practical range. Deeper beds give more contact time per square metre of footprint but add headloss and need a properly designed backwash system.

- **Hydraulic loading** — surface flow of **5–15 m/h** is typical. The higher the loading, the less contact time per pass, so EBCT and loading must be considered together.

## Mesh Size: 8x30 or 12x40?

**8x30** has larger particles: lower headloss and higher allowable flow, which suits large filters and retrofit projects with existing pumps. **12x40** packs more surface area per volume, giving faster kinetics — the standard choice for taste-and-odour control and deeper contactors. Both grades are common in drinking water; the right pick depends on your headloss budget, underdrains and flow profile. When in doubt, ask the supplier to model both against your flow rather than selling you one grade.

## Backwash: Where Hardness Pays

Backwash expands the bed by roughly **20–30%** to flush fines and re-stratify the media. Coconut-shell GAC routinely tests at **95–98% hardness**, so it sheds few fines over years of service. Softer coal-based media generates more fines that plug underdrains and add turbidity — a cost that shows up after the first backwash cycle, not on the quote. If your plant has aggressive backwashing, hardness is a buying criterion, not a footnote. See our coconut vs coal comparison (https://www.yelicarbon.com/blog-water-coconut-coal.html) for the full trade-off.

## When to Replace the Media

GAC is a consumable, and the clock starts at first contact. Plan for these triggers:

- **Iodine decay** — sample the bed and compare against the original COA. When iodine drops well below spec (many plants replace in the 600–800 mg/g range for organics duty), adsorption capacity is gone regardless of how the carbon looks.

- **Breakthrough before design EBCT** — if the target contaminant appears earlier than your design assumed, the bed is exhausted or channelized; check distribution, then plan change-out.

- **Fines and pressure drop** — rising differential pressure and turbidity after backwash point to media breakdown, not a flow problem.

Keep the original batch COA and re-test samples at intervals — a simple iodine or methylene blue test gives you an objective replacement signal. Our iodine vs methylene blue guide (https://www.yelicarbon.com/blog-iodine-test.html) explains which test fits your duty.

## What to Put in Your RFQ

Make the quote comparable between suppliers: grade (mesh 8x30 or 12x40), iodine 900–1100 mg/g, hardness ≥98% (or stated), moisture ≤5%, ash within grade, compliance documentation (NSF/ANSI 61 documents on request), and a **COA with every batch**. Then order a small trial batch before committing to container volume — a 1–2 m³ trial tells you more than any datasheet, and it validates the paperwork before you scale.

### Drinking-Water GAC, Documented Per Batch

Coconut GAC for municipal and industrial water — 8x30 / 12x40, iodine 900–1100 mg/g, hardness ≥98%, NSF/ANSI 61 documentation on request, COA with every batch. Factory-direct from Fujian, China.

Request a Water-Grade Quote → (https://www.yelicarbon.com/contact.html#quote-form)
Prefer WhatsApp? Chat with our water treatment team → (https://wa.me/8613600906330?text=Hello%20YELI%20Carbon!%20I%20would%20like%20a%20water-grade%20GAC%20quote.)

## The Bottom Line

Good carbon in a bad bed still fails. Specify EBCT, depth and loading with the same care as the carbon grade, monitor iodine and pressure drop from day one, and replace on data rather than calendar. For the complete buying process, see our municipal GAC buyer's checklist (https://www.yelicarbon.com/blog-post.html) and the PFAS design guide (https://www.yelicarbon.com/blog-pfas-gac.html).

## FAQ: GAC Bed Design for Drinking Water

What empty bed contact time should I design for GAC? +
For taste-and-odour compounds (geosmin/MIB) and moderate organics, plan for 7.5–15 minutes EBCT, treating 5 minutes as the absolute floor. For tougher duty such as PFAS, designs often use 15–20 minutes or more. Longer EBCT improves removal and extends bed life but requires bigger vessels.

What mesh size GAC should I use — 8x30 or 12x40? +
8x30 has larger particles: lower headloss and higher allowable flow, good for large filters and retrofit projects. 12x40 packs more surface area per volume for faster kinetics — the standard choice for taste-and-odour control and deeper contactors. Ask your supplier to model both against your flow rather than selling one grade.

When should I replace GAC media? +
Replace on data, not calendar: when sampled iodine drops well below spec (many plants replace in the 600–800 mg/g range for organics duty), when the target contaminant breaks through before design EBCT, or when differential pressure and turbidity rise from media breakdown. Keep the original batch COA and re-test at intervals.

What specs should a drinking-water GAC RFQ include? +
Grade (mesh 8x30 or 12x40), iodine 900–1100 mg/g, hardness ≥98% (or stated), moisture ≤5%, ash within grade, NSF/ANSI 61 documentation on request, and a COA with every batch. Order a 1–2 m³ trial batch before committing to container volume.

### Related reading

- Coconut Shell vs Coal-Based Activated Carbon for Drinking Water Treatment (https://www.yelicarbon.com/blog-water-coconut-coal.html)

- PFAS Removal with Granular Activated Carbon (https://www.yelicarbon.com/blog-pfas-gac.html)

- Granular Activated Carbon — Full Specification (https://www.yelicarbon.com/product-gac.html)

Browse all activated carbon products & request a quote → (https://www.yelicarbon.com/products.html)