Bulk Silo & Hopper Inventory Governance: 3D Radar Level Profiling, Janssen Compaction Density Correction, and Moisture-Adjusted Dry-Matter Mass Balance

Executive Summary: Why Single-Point Level Sensors Fail in Bulk Solids Silos

Unlike liquids in a tank—which self-level horizontally and maintain uniform incompressible density, dry bulk solids (grain, oilseeds, sugar, cement clinker, alumina, plastic pellets) exhibit internal friction, wall friction, compressibility, and hygroscopic moisture exchange:

1. Surface Topography Error (Angle of Repose $\alpha_r \approx 23^\circ\text{–}38^\circ$):
   • During Center Filling, the material forms an upward conical peak (Cone-Up). A single-point sensor aimed near the center overstates the true average bed height by up to $+1.5\text{ to } +3.0\text{ meters}$.
   • During Gravity Discharge (Core / Funnel Flow), the center collapses into an inverted conical crater (Cone-Down / Rat-Hole). That same single-point sensor now reads the bottom of the crater, understating inventory by $-15%$ or more, or—worse—hangs up on a bridged/ratted crust while the cavity underneath is empty.
2. Compaction Density Error (8% to 14% Overburden Compression):
   • Uncompacted loose wheat entering the top of a $30\text{-meter}$ silo has a test weight (bulk density) of $\rho_0 = 770\text{ kg/m}^3$. Under the overburden vertical pressure of $25\text{ meters}$ of grain above it, plus vibration from train/truck traffic and aeration fans, the effective average bulk density in the cylinder rises by a Compaction Factor (Pack Factor) of $k_{\text{pack}} = 1.055\text{ to } 1.095$ ($\bar{\rho}_{\text{eff}} \approx 820\text{ kg/m}^3$). Ignoring compaction causes a systematic 6%–9% book-to-physical mismatch.
3. Moisture Shrink vs. Physical Theft/Spillage Confusion:
   • When $10,000\text{ metric tons}$ of freshly harvested maize is received at $18.5%$ wet-basis moisture (MC_in) and dried/aerated down to the safe storage spec of $14.0%$ (MC_out), the total wet scale weight drops by $523.26\text{ metric tons}$ ($5.23%$)—even if zero kernels of dry matter were lost! Without a dual Wet-Tons + Dry-Matter (0% Moisture Equivalent) ERP ledger, finance controllers cannot tell whether a 550 MT silo shortage is normal evaporative water loss or 26.7 MT of physical grain theft/spillage.

Technical Architecture: 3D Radar Mesh, Compaction Engine, and Dual Wet/Dry ERP Ledger

Mathematical Foundation: 3D Conical Volume, Janssen Compaction, and Moisture Shrink

1. True 3D Silo Volume ($V_{\text{true}}$) Across Hopper, Cylinder, and Top Cone/Crater

For a cylindrical silo of internal radius $R$ ($D = 2R$) with a bottom conical discharge hopper of height $h_{\text{hopper}}$ and lower outlet radius $r_0$, where an 80 GHz 3D scanning radar (or multi-point phased array) measures the top surface height distribution $z(r, \theta)$ over the cross-section (or decomposes it into mean wall contact height $h_{\text{wall}}$ and upper positive/negative surface cone height $\Delta h_{\text{cone}} = \pm R \tan\alpha_r$, where $+$ is filling Cone-Up and $-$ is discharging Cone-Down):

$$V_{\text{hopper}} = \frac{\pi , h_{\text{hopper}}}{3}\left(R^2 + R r_0 + r_0^2\right)$$

$$V_{\text{true}} = V_{\text{hopper}} + \int_{0}^{2\pi}\int_{0}^{R} z(r, \theta), r, dr, d\theta \approx V_{\text{hopper}} + \pi R^2 h_{\text{wall}} + \frac{1}{3}\pi R^2 \Delta h_{\text{cone}}$$

Notice that the difference between a full Cone-Up peak ($+\frac{1}{3}\pi R^3 \tan\alpha_r$) and a full Cone-Down draw crater ($-\frac{1}{3}\pi R^3 \tan\alpha_r$) in a single $D = 18.0\text{ m}$ ($R = 9.0\text{ m}$) silo storing grain with $\alpha_r = 28^\circ$ ($\tan 28^\circ = 0.5317$) is:

$$\Delta V_{\text{topography}} = \frac{2}{3}\pi R^3 \tan\alpha_r = \frac{2}{3}\pi (9.0)^3 (0.5317) = 811.8\text{ m}^3 \approx 650\text{ metric tons of grain!}$$

2. Janssen Vertical Overburden Pressure & Compaction Pack Factor ($k_{\text{pack}}$)

By Janssen's differential slice equation (EN 1991-4), vertical stress $\sigma_v(z)$ at depth $z$ below the grain surface (where $K$ is the lateral-to-vertical stress ratio, $\mu_w$ is the coefficient of grain-on-wall friction, and $R_h = D / 4$ is the hydraulic radius of a circular silo) approaches an asymptotic limit as wall friction supports part of the column weight:

$$\sigma_v(z) = \frac{\rho_0 , g , R}{2 , K , \mu_w}\left(1 - e^{-\frac{2 K \mu_w z}{R}}\right)$$

Integrating compressible bulk density $\rho(z) = \rho_0 \left(1 + \beta_c , \sigma_v(z)^{\gamma}\right)$ over the filled height $H_{\text{eq}}$ yields the Effective Silo Pack Factor ($k_{\text{pack}}$) (ASABE W.A.M. / Pack-Factor Method):

$$\bar{\rho}{\text{eff}} = \rho_0 \cdot k{\text{pack}}(H_{\text{eq}}, D, \text{MC}, \text{WallType}), \quad M_{\text{physical_wet}} = V_{\text{true}} \cdot \bar{\rho}_{\text{eff}}$$

3. Exact Evaporative Moisture Shrink vs. Handling Dry-Matter Shrink

When a wet lot of mass $M_{\text{in}}$ entering at wet-basis moisture fraction $m_{\text{in}} \in (0, 1)$ (e.g., 0.185 for 18.5%) is conditioned/dried to target moisture fraction $m_{\text{out}}$ (e.g., 0.140 for 14.0%), conservation of dry matter $M_{\text{DM}} = M_{\text{in}}(1 - m_{\text{in}}) = M_{\text{out_expected}}(1 - m_{\text{out}})$ dictates the exact Evaporative Water Weight Shrink ($\Delta M_{\text{water}}$):

$$S_{\text{moisture_pct}} = \left(\frac{m_{\text{in}} - m_{\text{out}}}{1 - m_{\text{out}}}\right) \times 100%, \quad M_{\text{out_expected}} = M_{\text{in}} \cdot \left(\frac{1 - m_{\text{in}}}{1 - m_{\text{out}}}\right)$$

*(Note: A common industry blunder is subtracting $18.5% - 14.0% = 4.50%$ instead of dividing by $(1 - 0.140) = 0.860$, which yields the true mass reduction of $5.2326%$!)*

Any residual gap between $M_{\text{out_expected}}$ and actual measured physical mass $M_{\text{physical_wet}}$ is True Dry-Matter / Handling Variance ($\Delta M_{\text{handling}}$) (dust aspiration, respiration, spillage, or theft):

$$\Delta M_{\text{handling_wet_eq}} = M_{\text{physical_wet}} - M_{\text{in}} \cdot \left(\frac{1 - m_{\text{in}}}{1 - m_{\text{actual}}}\right)$$

Worked Numerical Example: 18m Concrete Grain Silo (Silo #07 — Hard Red Winter Wheat)

Consider Silo #07 (slip-formed concrete cylinder, internal diameter $D = 18.0\text{ m} \implies R = 9.0\text{ m}$, cylinder height $H_{\text{cyl}} = 34.0\text{ m}$, bottom $40^\circ$ steel discharge hopper height $h_{\text{hopper}} = 7.20\text{ m}$, outlet radius $r_0 = 0.40\text{ m}$):

1. Cumulative Inbound Receipts & Conditioning:

   • Across the harvest campaign, Weighbridge + Automated Probe Sampler records total inbound receipts into Silo #07 of $M_{\text{in}} = 6,800.00\text{ MT}$ at weighted-average inbound moisture $m_{\text{in}} = 16.20%$ and uncompacted laboratory test weight $\rho_0 = 0.775\text{ MT/m}^3$ ($775\text{ kg/m}^3$).
   • In-silo drying/aeration brings the current grain column moisture (measured by multi-sensor digital moisture/temperature cables) to $m_{\text{actual}} = 13.50%$.
   • Expected Moisture-Adjusted Book Mass ($M_{\text{book_expected}}$ at $13.50%$ MC): $$S_{\text{moisture_pct}} = \frac{0.1620 - 0.1350}{1 - 0.1350} = \frac{0.0270}{0.8650} = 3.1214% \quad (\Delta M_{\text{water}} = -212.25\text{ MT evaporated water})$$ $$M_{\text{book_expected}} = 6,800.00 \times (1 - 0.031214) = \mathbf{6,587.75\text{ MT}} \quad (\text{Dry Matter } M_{\text{DM}} = 5,698.40\text{ MT})$$

2. 3D Radar Surface Scan & Compaction Audit after Partial Outload Draw (Cone-Down):

   • No commercial outload has been billed yet, wait—let's evaluate the silo after a $500.00\text{ MT}$ internal transfer outload (leaving expected book mass = 6,587.75 - 500.00 = 6,087.75 MT at $13.50%$ moisture, or $M_{\text{DM, book}} = 5,265.90\text{ MT}$):
   • Because the $500\text{ MT}$ transfer was drawn from the bottom center gate, the grain top surface is now an inverted Cone-Down crater with wall contact height $h_{\text{wall}} = 28.40\text{ m}$ above the hopper transition and center depression $\Delta h_{\text{cone}} = -4.50\text{ m}$.
   • Hopper Volume ($V_{\text{hopper}}$): $$V_{\text{hopper}} = \frac{\pi (7.20)}{3}\left(9.0^2 + 9.0(0.40) + 0.40^2\right) = 2.4\pi \times 84.76 = 639.08\text{ m}^3$$
   • Cylinder + Inverted Crater Volume ($V_{\text{cyl+cone}}$): $$V_{\text{cyl+cone}} = \pi (9.0)^2 (28.40) + \frac{1}{3}\pi (9.0)^2 (-4.50) = 7,226.92 - 381.70 = 6,845.22\text{ m}^3$$
   • True 3D Solid Volume ($V_{\text{true}}$): $$V_{\text{true}} = 639.08 + 6,845.22 = \mathbf{7,484.30\text{ m}^3}$$
   • (Note: A legacy single-point center radar pointing into the $-4.50\text{ m}$ crater would assume a flat cylinder at $h_{\text{center}} = 23.90\text{ m}$, calculating $V_{\text{legacy}} = 639.08 + \pi(81)(23.90) = 6,720.89\text{ m}^3$—missing $763.41\text{ m}^3$!)
   • Janssen Compaction Effective Bulk Density ($\bar{\rho}_{\text{eff}}$): At equivalent depth $H_{\text{eq}} = 26.9\text{ m}$ in a $D = 18.0\text{ m}$ concrete silo ($\mu_w = 0.40, K = 0.52$), the ASABE compaction factor is $k_{\text{pack}} = 1.0480$: $$\bar{\rho}_{\text{eff}} = 0.775 \times 1.0480 = \mathbf{0.8122\text{ MT/m}^3}$$
   • Physical Audited Wet Mass ($M_{\text{physical_wet}}$): $$M_{\text{physical_wet}} = 7,484.30\text{ m}^3 \times 0.8122\text{ MT/m}^3 = \mathbf{6,078.75\text{ MT}}$$
   • Physical-to-Book Reconciliation: $$\Delta M_{\text{handling}} = 6,078.75\text{ MT} - 6,087.75\text{ MT} = -9.00\text{ MT} \quad (-0.148%\text{ dust/handling loss vs. } 1.25%\text{ audit threshold } \implies \mathbf{PASS})$$

Silo Telemetry & ERP Mass-Balance Reconciliation Payload

{
  "siloInventoryAuditRecord": {
    "auditId": "SILO-AUD-2026-10-20-S07",
    "terminalId": "PORT-GRAIN-TERM-02",
    "siloId": "SILO_07",
    "commodityCode": "WHEAT-HRW-NO2",
    "geometrySpec": {
      "diameterM": 18.0,
      "cylinderHeightM": 34.0,
      "hopperHeightM": 7.2,
      "wallMaterial": "SLIPFORM_CONCRETE"
    },
    "radar3dScan": {
      "sensorModel": "80GHZ_3D_PHASED_ARRAY",
      "surfaceState": "CONE_DOWN_DISCHARGE",
      "wallContactHeightM": 28.4,
      "centerApexDeltaM": -4.5,
      "hopperVolumeM3": 639.08,
      "cylinderAndConeVolumeM3": 6845.22,
      "totalTrueVolumeM3": 7484.30
    },
    "densityAndCompaction": {
      "labUncompactedTestWeightMtPerM3": 0.775,
      "janssenPackFactor": 1.048,
      "effectiveCompactedDensityMtPerM3": 0.8122,
      "physicalAuditedWetMassMt": 6078.75
    },
    "moistureAndDryMatterLedger": {
      "weightedAvgInboundMoisturePct": 16.2,
      "currentCableAvgMoisturePct": 13.5,
      "cumulativeEvaporatedWaterShrinkMt": 212.25,
      "moistureAdjustedBookWetMassMt": 6087.75,
      "bookDryMatterMassMt": 5265.9,
      "physicalDryMatterMassMt": 5258.12,
      "netHandlingVarianceWetEqMt": -9.0,
      "netHandlingVariancePct": -0.148,
      "auditStatus": "RECONCILED_PASS"
    }
  }
}

Verification & Acceptance Playbook (ACT-L159)

1. ACT-L159-01 — Evaporative Moisture Shrink vs. Dry-Matter Conservation Gate:
   • Scenario: Post an inbound receipt of 1,000.00 MT of wet corn at 19.0% moisture into Dryer/Silo S-12, dry the lot to 14.0% moisture, and weigh out the entire lot at 941.86 MT (exact theoretical water loss = 58.14 MT).
   • Expected System Behavior: The ERP silo inventory management module must record 58.14 MT (5.814%) against the EVAPORATIVE_MOISTURE_SHRINK ledger account and 0.00 MT (0.00%) against UNEXPLAINED_DRY_MATTER_VARIANCE—never misclassifying evaporated water as physical inventory shrinkage.
2. ACT-L159-02 — Bridging / Rat-Hole Cavity Detection Interlock:
   • Scenario: Open the bottom discharge slide gate of Silo #04 and record 180 MT moving across the outbound belt scale while the 3D radar surface scanner detects 0.0 m³ change in the upper surface elevation (grain has bridged/crusted across the top while a hollow cavity forms underneath).
   • Expected System Behavior: Within 15 minutes (or 50 MT of unconfirmed belt-scale draw without corresponding upper surface drop), the PLC/WCS must close the silo discharge slide gate, raise CRITICAL_ALARM_SILO_BRIDGE_CAVITY, and prohibit personnel entry (preventing catastrophic bridge collapse or structural implosion).
3. ACT-L159-03 — Automated Distributor Valve Interlock Against Cross-Grade Bin Contamination:
   • Scenario: Create an inbound dump ticket for GMO Yellow Maize on Receiving Pit #2 while the overhead rotary distributor spout is physically positioned over Silo #02 (currently holding Non-GMO Identity-Preserved Food-Grade Maize).
   • Expected System Behavior: The PLC interlock must check the inductive proximity switch on the rotary distributor spout against the ERP Silo Grade Matrix, refuse to start the bucket elevator (ERR_GRADE_CONTAMINATION_INTERLOCK), and hold Pit #2 until the distributor locks onto an approved GMO Yellow Maize bin.

Frequently Asked Questions (FAQ)

When should a facility use load cells versus 3D radar scanners for silo inventory?

• Gravimetric Load Cells (under hopper skirts or legs) are the gold standard for small-to-medium steel hoppers (50 MT to 500 MT) because they measure mass directly regardless of angle of repose or compaction. However, on large flat-bottom concrete or bolted-steel storage silos (2,000 MT to 15,000 MT sitting on concrete ring foundations), load cells cannot be retrofitted underneath the floor, and wind-load/thermal expansion on tall outdoor structures degrades strain-gauge accuracy. For large silos, 80 GHz 3D Radar Surface Profiling + Janssen Compaction + Belt-Scale Mass Balance is the industry standard.

Why do grain terminals track both "Wet Metric Tons" and "Standard Moisture Basis Tons (e.g., 14.0%)"?

Commercial grain contracts specify a base moisture (e.g., 14.0% for wheat/corn, 13.0% for soybeans). If a farmer delivers corn at 17.5% moisture, the terminal cannot pay wheat/corn price for the extra 3.5% water—nor can it sell that water to an export vessel. Converting every receipt, storage bin, and blend into Standard Moisture Equivalent Tons (14.0% basis) and Bone-Dry Matter (0.0% basis) ensures the commercial desk hedges and blends exact dry-matter mass.

How do in-silo temperature and moisture cables prevent spontaneous heating and spoilage?

Biological respiration of grain and mold growth generate heat and moisture. By hanging multi-sensor digital temperature/RH cables every 4–6 meters radially inside the silo, the silo inventory management system calculates the Rate of Temperature Rise ($\Delta T / \Delta t > +1.5^\circ\text{C per 48 hours}$) at each 3D voxel—automatically turning on variable-speed aeration fans during low-humidity ambient night windows or scheduling a "bin turn" before a hot spot chars or cakes the grain.

How does dust collection (baghouse aspiration) affect the silo mass balance?

Environmental air permits require negative-pressure dust aspiration at every conveyor transfer point, bucket elevator boot/head, and silo fill spout. Typical dust collection extracts 0.10% to 0.25% of total throughput weight into the baghouse dust bin. Weighing the collected baghouse dust hoppers and crediting that weight against the terminal mass balance accounts for up to half of residual handling shrink.