Hempnall with low flow
You’re absolutely right to challenge that. If a supplier draws a hard line at ~5 L/s DWF, it’s not marketing theatre, it’s physics, hydraulics, and a bit of commercial scar tissue from plants that didn’t behave. My earlier confidence wasn’t misplaced on process capability, but it didn’t fully respect the scale-dependent hydraulics. Let’s unpack that properly.
🔍 Why suppliers get nervous below ~5 L/s DWF
At small flows, tertiary denitrifying filters stop behaving like elegant polishing units and start acting like temperamental houseplants 🌱
1. Hydraulic distribution breaks down
Denitrifying sand filters rely on:
Even flow distribution across the bed
Stable loading rates
At low DWF (~<5 L/s):
Flow becomes intermittent or channelised
Parts of the bed go:
Starved (no substrate)
Or overfed (short-circuiting)
👉 Result: inconsistent denitrification, nitrate leakage
2. Backwash becomes disproportionate
Backwashing is not optional. It’s the heartbeat.
At small works:
Backwash volume = large % of daily flow
You get:
Hydraulic shocks upstream
Process instability
High recycle fractions
👉 The plant starts chasing its own tail
3. Carbon dosing control gets twitchy
At low flows:
Small dosing errors = big concentration swings
Online control becomes noisy
👉 You oscillate between:
Under-dosing → nitrate breakthrough
Over-dosing → residual COD, permit risk
4. Minimum viable reactor size
There’s a practical lower limit where:
Media depth
Contact time
Underdrain hydraulics
…just don’t scale down gracefully.
👉 You can build it smaller. It just won’t behave like the big ones.
⚖️ So was the original recommendation wrong?
Not fundamentally, but it needs refinement:
Technically valid: Yes, tertiary filters can achieve the TN target
Practically appropriate at 7 L/s FFT (~3–5 L/s DWF)? Often no
Supplier reluctance? Completely justified
Think of it like specifying a racehorse for a village lane. It’s capable, just not suited to the terrain.
🧭 What I would recommend instead (adjusted for reality)
Now we lean into your strengths:
👉 small works, cold temp, plenty of space
🥇 Revised Best Option: Oversized MLE with Step-Feed + Carbon Dosing (No Tertiary)
Why this works better at small scale
Fully biological, no fragile tertiary hydraulics
Continuous process → stable at low flows
Uses space to compensate for cold temperature
Key design features
1. Step-feed configuration
Instead of one anoxic zone:
Split influent across 2–3 anoxic stages
👉 This:
Spreads carbon availability
Improves denitrification efficiency
Reduces reliance on perfect mixing
2. Oversized anoxic volume
Push anoxic fraction toward 35–45%
Increase HRT
👉 You’re buying robustness with concrete instead of chemicals
3. External carbon (strategic dosing)
Dose into later anoxic stages, not just front-end
Trim nitrate “tail”
4. High SRT (critical at 8°C)
Target 20–25 days
Ensure nitrification never becomes limiting
Expected performance
Consistent TN: 8–12 mg/L
With tuning: reliably ≤10 mg/L
🥈 Alternative: MLE + IFAS (if you want insurance)
If you’re nervous about winter nitrification:
Add carriers in aerobic zone
Boost nitrifier population
👉 Particularly useful if:
Influent ammonia is variable
Future tightening is likely
🥉 Niche option: Package continuous denitrification reactors
Some suppliers offer:
Small-scale continuous denitrification units (not sand filters)
These:
Avoid backwash issues
Are better suited to low flows
👉 Worth exploring, but vendor-dependent
🚫 What I would now avoid
❌ Denitrifying sand filters (at this scale)
❌ SBR (unless operator skill is high)
❌ MBR (overkill for this consent)
🧠 The deeper principle (this is the real takeaway)
At small flows:
Hydraulics dominate over kinetics
At large works:
Biology is the limiting factor
At small works:
Flow stability, distribution, and control dominate
That’s why:
Big plants love tertiary filters
Small plants quietly suffer with them
🧾 Final position (engineer-to-engineer)
You were right to question the recommendation.
Given:
7 L/s FFT
Likely ~3–5 L/s DWF
8°C
Available land
👉 The most buildable, operable, and reliable solution is:
✅ Large, staged MLE with step-feed + targeted carbon dosing
No fragile polishing step.
No hydraulic gymnastics.
Just a system that behaves the same way on a quiet Tuesday as it does during a storm.
If you’d like, I can:
Sketch a process flow diagram
Or size anoxic/aeration volumes and recycle ratios for a UK-standard influent
We can turn this from concept into something you could actually take to outline design.
Comments
Post a Comment