Investigating Fermentation of Primary Sludge in Full-Scale Gravity Thickeners

Introduction
Wastewater treatment plants face stricter nitrogen (TN) discharge limits, increasing reliance on external carbon (C) sources for nitrogen (N) removal, which drives up costs and external dependencies. Advances in short-cut N removal processes, like PdNA, reduce carbon use but still depend on significant external C dosing. Influent wastewater carbon is less effective for N removal due to aerobic consumption, whereas external carbon dosing avoids oxygen-related losses (Ladipo-Obasa et al., 2022). Primary sludge fermentate is a low-cost carbon source that could reduce reliance on external carbon, such as methanol (MeOH) (Ali et al., 2021). However, variable organic yields and fermentate quality present challenges (Sánchez Rubal et al., 2012). Factors like wastewater characteristics, seasonal conditions, and sewer fermentation activity affect consistency.

Preliminary bench-scale tests at Blue Plains showed potential yields of 0.2 g sCOD/g VSS but also increased phosphorus release and odor. This study expands fermentation tests in full-scale gravity thickeners to assess fermentate quality, nutrient release, and odor formation under realistic conditions, optimizing N and P removal processes.

Materials and Methodology
Fermentation tests were conducted in the gravity thickener (GT) at Blue Plains AWWTP over seven months with controlled sludge blanket height covering a range of SRTs and varying operational and seasonal conditions (Fig. 1). Grab samples were collected from the primary sludge influent, GT overflow, and three depths of sludge blanket. A complete elutriation assumption was applied to estimate weighted organic yields (sCOD), nutrient release (orthophosphate, ammonia, iron). Additionally, odor samples were tested for odor detection thresholds (ASTM E679 and EN13725) and total reduced sulfur and compositions (ASTM D5504-20), while a gas analyzer (Odalog) provided real-time H2S monitoring onsite.

Result and Discussion
SRT control in gravity thickeners
SRT in gravity thickeners was controlled by maintaining sludge blanket targets (Fig. 1A). Flow to the gravity thickener was stabilized to enhance consistency (Fig. 1B). Improved system stability typically led to higher sCOD yields, while instability, often during morning samples, resulted in lower yields (Fig. 1C, Fig. 1D).

Stability, assessed via carbon (C) and phosphorus (P) mass balance errors, was higher when errors stayed below 20%. Yields varied within the 1—1.5-day SRT range, with better stability crucial for optimizing fermentate quality in full-scale operations.

COD and nutrient yields (N, P, Fe) under full elutriation assumption
The COD and nutrient yields were based on full gravity thickener profiles, reflecting 100% elutriation since no elutriation occurred under current conditions. Typically, 20% of soluble compounds from fermentation are lost in the underflow (WRF Project 4975, 2023). sCOD yield increased with SRT (Fig. 2A), though variability beyond 1 day suggests other influencing factors. Among 11 data points (SRT 1—1.5 days), five showed high-quality fermentate (>0.2 g sCOD/g VSS), while six showed lower quality (<0.2 g sCOD/g VSS). High-quality fermentate correlated with mass balance errors <30%, while poor quality correlated with >30% errors, indicating instability in GT operations. Factors such as primary clarifier operation and GT underflow influenced stability, with lack of control strategies being a bottleneck. NH4-N yield varied without clear SRT trends (Fig. 2B), averaging 20 ± 7.5 mg N/L at SRT >1 day. P release increased with SRT (Fig. 2C), with 1.64 g Fe released per gram of OP-P, enabling 61% P removal using current Fe/P ratios.

GTO and future fermentate composition
Previous studies showed that increased SRT improved compression (SVI30 from 48 to 36 mL/g) but created smaller, harder-to-flocculate particles (TOF > 700 mg TSS/L) (Islam et al., 2024). Higher sludge blankets at SRT > 1.25 days raised GTO TSS (Fig. 2E). To balance sCOD and nutrient yields with settling, too long SRTs should be avoided.

In Table 1, GTO and thus fermentate composition at a SRT of 1-1.25 days are calculated assuming 80% elutriation. This can lead to current GTO flows and denitrification stoichiometry and yield coefficient of 5.4 sCOD/g NO3-N (Lee et al., 2024) to a potential MeOH saving of 22±8%. When using fermentate for PdNA, this MeOH saving can be increased to 28 to 62% at PdNA efficiencies of 30-60%.

Odor emissions
Odor is a key side effect of fermentation, with H2S measured continuously and odor units/organic sulfur assessed at intervals. H2S levels reached 8537 ppbv, and odor detection peaked at 360,000 OU/m3 for SRT <1.25 days. Odor emissions increased in level and variability at SRT >1.5 days (Fig. 3), correlating with higher SRTs. Maintaining an optimal SRT (1—1.25 days) minimizes excessive microbial degradation and odor emissions while preserving fermentate quality. Further investigation is needed to understand variability and improve odor treatment design without overestimating for worst-case scenarios.

Conclusion
This study quantified fermentation yields, nutrient release, and odor production during full-scale gravity thickening under varying conditions. It provides guidelines to optimize fermentate quality with high sCOD yield, advancing fermentation to reduce MeOH use and leveraging GT as a fermenter.