Appendix

Drawings gallery · P&ID · psychrometric chart · mass balance · sensitivity · calculations · glossary

Drawings gallery — fabrication drawings

S2 — Longitudinal section (supply channel)

Drawings gallery — process & control diagrams

S7 — P&ID: refrigerant + air + condensate

S8 — Psychrometric chart with state points

3D renders

Validation results

The thermodynamic model in the source tool is cross-checked against an independent Python implementation at every commit. For the target configuration, all 20 compared values match to 3 decimal places (relative error < 1%):

Quantity	Tool	Reference	Status
Tray area	181.9 m²	181.9	OK
Fresh batch	1,819 kg	1,819	OK
Drying time	19.7 h	19.7	OK
Airflow	10,653 m³/h	10,653	OK
HP required	85.9 kW	85.9	OK
SMER	4.03	4.03	OK
Pressure budget	410 Pa	410	OK

…13 more values all match. Sanity ranges pass: drying 10–20 h, SMER 2.5–4.5 kg/kWh, coil > 5°C (no frost), return RH < 90%.

Mass & energy balance (per batch)

Flow	In	Out	Notes
Fresh coconut (50% MC)	1,819 kg	—	Loaded on trays
Water removed	—	862 kg	Condensed at evaporator, drained
Copra out (5.0% MC)	—	957 kg	Unloaded after cycle
Solids (dry matter, conserved)	910 kg	895 kg	Check: solids in = solids out ✓

Energy flow	kWh/batch	Notes
Latent heat (water evaporated)	~570	Recovered at evaporator, reused at condenser
Structure heat loss (walls)	~25	Lost to ambient through PUF panels
Product heat-up	~41	30°C → 62°C, transient (first 1.5 h)
Fan motor heat	~34	Enters air stream (offsets condenser)
Total electricity consumed	214	= (thermal / COP) + fan

Sensitivity analysis

How robust is the design to variation in key inputs? Each row changes one parameter by ±15–20% from the design point and reports the impact.

Parameter	Low	Design	High	Impact
Loading (kg/m²)	8	10	12	Batch ±20%; SMER stable 3.4–3.7; drying time unchanged (ratio-based)
Supply temp (°C)	57	62	65	Drying time 24 h → 19.7 h → 16 h; case-hardening risk rises above 65°C
Face velocity (m/s)	2.4	2.8	3.2	Return RH 85→71→60%; fan power +44%; ΔP rises proportionally
Ambient temp (°C)	25	30	35	Wall loss ±0.4 kW (negligible); HP sizing unaffected
HP COP	2.8	3.2	3.6	SMER 3.2→3.6→4.0; electricity per batch ±12%

All sensitivities computed within the tool by adjusting the relevant parameter and re-running the thermo engine. Design point is highlighted; the system remains feasible (no warnings) across the full range shown.

Worked calculations summary

HP sizing (how 86 kW is derived)

peak_rate = water / t_dry × peak_factor = 862 / 19.7 × 1.3 = 56.8 kg/h
peak_latent = peak_rate × h_fg / 3600 = 56.8 × 2381 / 3600 = 38 kW
required = (peak_latent + loss + product_heatup) × 1.15
= (38 + 1.7 + 28.7) × 1.15 = 86 kW
installed = 2 × 50 kW modules (staged)

Wall U-value

U = 1 / (R_inside + R_PUF + R_outside)
= 1 / (0.13 + 0.050/0.022 + 0.04)
= 1 / 2.44 = 0.41 W/m²·K
loss = U × A_env × ΔT / 1000 = 0.41 × 92.8 × 32 / 1000 = 1.2 kW

Fan affinity law (why velocity matters)

Q ∝ N (flow ∝ speed)
ΔP ∝ N² (pressure ∝ speed²)
Power ∝ N³ (power ∝ speed³)
→ A 15% velocity increase costs 52% more fan power.

Psychrometric formulas

p_ws(T) = 0.61094 · exp(17.625·T / (T + 243.04)) kPa  [Magnus, water]
W(T, RH) = 0.62198 · p_w / (P − p_w)  where p_w = RH · p_ws(T)
h(T, W) = 1.006·T + W·(2501 + 1.86·T)  kJ/kg_da
v_da(T, W) = 287.055 · (T + 273.15) / ((P − p_w) · 1000)  m³/kg_da
T_dp(W) = inverse Magnus on p_w = W·P / (0.62198 + W)

Glossary

SMER	Specific Moisture Extraction Rate — kg of water removed per kWh of electricity consumed
COP	Coefficient of Performance — heat delivered / electrical input (heat-pump basis)
W (humidity ratio)	kg of water vapour per kg of dry air
Page model	Thin-layer drying equation MR = exp(−k·tⁿ)
PUF	Polyurethane Foam (insulation); 50 mm sandwich panel
U-flow	Air path: supply down one channel, U-turn at door end, return via parallel channel
Circuit	One 1-U-turn air path (this design: 1 circuit, 2 channels)
Abreast	Trolleys side-by-side within a channel sharing the same airflow